CN102955829B - For the method being ranked up to resource items, device and equipment - Google Patents

Abstract

The present invention provides a method, device and equipment for sorting resource items. According to the scheme of the present invention, the timeliness information related to the input sequence is obtained according to the input sequence from the user; and according to whether the timeliness information meets the predetermined condition, judging whether the input sequence is time-sensitive; and when judging that the input sequence is time-sensitive, sorting the resource items retrieved based on the input sequence according to timeliness, so as to provide the sorted resource items to said user. Compared with the prior art, the present invention has the following advantages: 1) Time-sensitive sorting is performed on the resource items retrieved based on the time-sensitive input sequence, so as to advance the sorting of resource items that may be related to the time-sensitive input sequence , to better meet the user's sudden retrieval needs; 2) perform time-sensitive sorting according to the freshness of resource items, so as to obtain better time-sensitive sorting results.

Description

Method, apparatus and device for sorting resource items

technical field

The present invention relates to the field of computer technology, in particular to a method, device and equipment for sorting resource items.

Background technique

In the prior art, sudden retrieval requirements related to some unexpected events are often generated due to these unexpected events. This kind of demand has the characteristics of unpredictable, short-term burst, and concentrated retrieval demand. Therefore, there is still a lack of solutions in the prior art that can provide users with time-sensitive sorting results according to sudden retrieval demand.

Contents of the invention

The object of the present invention is to provide a method, device and equipment for sorting resource items.

According to one aspect of the present invention, there is provided a method for sorting resource items, wherein the method includes the following steps:

a Acquiring time-sensitive information related to the input sequence according to the input sequence from the user;

b. Judging whether the input sequence has timeliness according to whether the timeliness information satisfies a predetermined condition;

Wherein, the method also includes the following steps:

x When it is judged that the input sequence has timeliness, perform timeliness sorting on the resource items retrieved based on the input sequence, so as to provide the sorted resource items to the user.

According to another aspect of the present invention, a sorting device for sorting resource items is also provided, wherein the sorting device includes:

A timeliness information acquisition device, configured to acquire timeliness information related to the input sequence according to the input sequence from the user;

The first judging means is used for judging whether the input sequence has timeliness according to whether the timeliness information satisfies a predetermined condition;

The timeliness sorting device is configured to perform timeliness sorting on the resource items retrieved based on the input sequence when it is judged that the input sequence is timeliness, so as to provide the sorted resource items to the user.

Compared with the prior art, the present invention has the following advantages: 1) It can identify the time-sensitive input sequence, and can perform time-sensitive sorting on the resource items retrieved based on the time-sensitive input sequence, so as to sort the resource items that may be related to The timeliness-related resource items of the input sequence are sorted in advance, which better meets the user's sudden retrieval needs; 2) The timeliness sorting is performed according to the freshness of the resource items, so as to obtain better timeliness sorting results.

Description of drawings

Other characteristics, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

Fig. 1 is a flowchart of a method for sorting resource items according to one aspect of the present invention;

Fig. 2 is a flowchart of a method for sorting resource items according to a preferred embodiment of the present invention;

3 is a schematic structural diagram of a sorting device for sorting resource items according to one aspect of the present invention;

Fig. 4 is a schematic structural diagram of a sorting device for sorting resource items according to a preferred embodiment of the present invention.

The same or similar reference numerals in the drawings represent the same or similar components.

detailed description

The present invention will be described in further detail below in conjunction with the accompanying drawings.

Fig. 1 is a flowchart of a method for sorting resource items according to one aspect of the present invention. Among them, the method according to the present invention is mainly completed by the sorting device in the computer equipment, which includes but not limited to devices such as application modules, operating systems, processing controllers, etc. installed in the computer equipment and capable of implementing the solutions of the present invention . Wherein, the computer equipment includes but not limited to: 1) user equipment; 2) network equipment. The user equipment includes but is not limited to computers, smart phones, PDAs, etc.; the network equipment includes but is not limited to a single network server, a server group composed of multiple network servers, or a large number of computers or networks based on cloud computing (Cloud Computing). A cloud composed of servers, among them, cloud computing is a kind of distributed computing, a super virtual computer composed of a group of loosely coupled computer sets.

In step S1, the sorting device acquires timeliness information related to the input sequence according to the input sequence from the user.

Specifically, the sorting device obtains an input sequence provided by other devices or devices, or obtains an input sequence from a user in real time, and obtains time-sensitive information related to the input sequence according to the input sequence. Among them, the way for the sorting device to obtain the timeliness information related to the input sequence includes but is not limited to:

1) The sorting device performs a query in the pre-established user demand sudden increase keyword database according to the input sequence, and uses the query result obtained from the query as the time-sensitive information. Wherein, the query result includes any of the following items: a) It is used to indicate that the input sequence or the keywords contained in the input sequence can be obtained by querying the user demand sudden keyword database, for example, "Y", " 1" etc.; b) is used to indicate that the input sequence or the identification information of the keywords contained in the input sequence cannot be obtained from the user demand sudden keyword database query, for example, "N", "O" and so on. Preferably, the query result also includes at least one of the following: a) the sudden increase keyword in the user demand sudden increase keyword library obtained by querying the user demand sudden increase keyword database; Keyword-related information, for example, input sequence or user demand change intensity of keywords, etc.

2) The sorting device acquires time-sensitive characteristic information related to the input sequence according to the input sequence as the time-sensitive information. Wherein, the time-sensitive feature information includes feature information that can be used to correlate with the input sequence and can be used to judge whether the input sequence is a time-sensitive input sequence. Preferably, the time-sensitive feature information includes but not limited to at least the following item:

a) information on the change in the number of resource items over time obtained by searching based on the input sequence; for example, the information on the change in the number of resource items over time includes the following information: the resource items obtained by searching based on the input sequence contain 50 Resource items created within the previous 96 to 120 hours, 48 resource items created within the previous 72 to 96 hours, 127 resource items created within the previous 48 hours to the previous 72 hours, 309 resource items created within the previous 24 to Resource items created within the previous 48 hours and 1045 resource items created within 24 hours; as another example, the information on the change in the number of resource items over time includes the relative quantity of resource information created within each predetermined time period. The change value of the amount of resource information established in the adjacent time period, etc. Preferably, the sorting device obtains information on changes in the number of resource items over time by statistically analyzing the creation or update time of the resource items retrieved based on the input sequence.

b) information on the number of times the input sequence and/or keywords included in the input sequence are retrieved over time; for example, the input sequence or the number of times the keywords included in the input sequence are retrieved within each predetermined time period or at a predetermined time point ; Another example is the difference between each predetermined time period or predetermined time point relative to its adjacent time period or time point's input sequence or the number of times keywords are retrieved. Preferably, the sorting device performs a query in the local storage device based on the input sequence or keywords contained in the input sequence, so as to obtain pre-stored information on changes in the number of searches over time corresponding to the input sequence or keywords contained in the input sequence; Or, the sorting device sends a request to other devices or equipment to obtain information on the number of searches over time corresponding to the input sequence or keywords contained in the input sequence; The search times of the keywords contained in the sequence are monitored and counted to obtain the change information of the search times over time.

c) Semantic burst feature information of the input sequence obtained based on semantic analysis of the input sequence. Wherein, the burst semantic feature information includes burst feature keywords and/or burst feature semantic structure. For example, the burst semantic feature information includes time keywords such as "latest, today, this week, recent, certain year, month, and day", or includes time keywords such as "earthquake, car accident, death, fire, hacking incident, suicide door" and other emergency keywords; for another example, the emergency semantic feature structure includes structures such as proper nouns + common nouns; for another example, the emergency semantic feature information includes such as "Fukushima earthquake" and other emergency feature keywords and Semantic feature information of burst feature semantic structure. Preferably, preferably, the sorting device performs a query in the local storage device based on the input sequence to obtain pre-stored burst semantic feature information corresponding to the input sequence; or, the sorting device sends a request to other devices or devices to obtain The burst semantic feature information corresponding to the input sequence; or, the sorting device performs semantic analysis on the input sequence to obtain the burst semantic feature information of the input sequence.

3) The sorting device first inquires the sudden increase keyword related to the input sequence in the pre-established user demand sudden increase keyword library according to the input sequence; when the sudden increase keyword related to the input sequence is found, it will The query result is used as the timeliness information; when the sudden increase keyword related to the input sequence cannot be found, the sorting device obtains the timeliness feature information related to the input sequence according to the input sequence as the timeliness information A step of. Wherein, the manner in which the sorting device obtains the time-sensitive feature information related to the input sequence according to the input sequence is the same as or similar to the manner in which the sorting device obtains the time-sensitive feature information in 2) above, and will not be repeated here.

It should be noted that the above examples are only to better illustrate the technical solution of the present invention, rather than limit the present invention. Those skilled in the art should understand that any time-sensitive information related to the input sequence obtained according to the input sequence from the user All implementation methods should be included within the scope of the present invention.

Next, in step S2, the sorting device judges whether the input sequence has timeliness according to whether the timeliness information satisfies a predetermined condition. Wherein, the predetermined conditions include the conditions that need to be satisfied by the timeliness information obtained based on the input sequence that can be determined to be timeliness; When the query results are obtained, the predetermined conditions include identification information such as "Y" or "1" in the query results, and when the timeliness information includes timeliness feature information, the predetermined conditions include conditions that the timeliness feature information needs to meet, etc. Preferably, based on the input The faster the number of resource items retrieved by the sequence increases with time, the faster the number of retrievals of the input sequence or keywords contained in the input sequence increases with time, and the stronger the burst semantic feature contained in the burst semantic feature information , the sorting device is more likely to judge the input sequence as a time-sensitive input sequence.

For example, the predetermined condition includes that the query result contains the identification information "Y" or "1". In step S1, the timeliness information obtained by the sorting device based on the input sequence includes the query result containing the identification information "Y"; then the sorting device according to the timeliness The information satisfies the predetermined condition, and it is judged that the input sequence has timeliness.

For another example, the predetermined condition includes that the growth rate of the number of resource items established in any time period relative to the number of resource items established in the previous time period exceeds 50%; in step S1, the sorting device retrieves based on the input sequence to obtain The obtained change information of the number of resource items over time includes the following information: among the resource items retrieved based on the input sequence, there are 50 resource items established within the previous 96 to 120 hours, 48 resource items established within the previous 72 to Resource items created within 96 hours, 127 resource items created within the previous 48 hours to the previous 72 hours, 309 resource items created within the previous 24 to 48 hours, and 1045 resource items created within 24 hours; Then the sorting device analyzes the change information of the resource item quantity over time, determines the growth rate of the resource item quantity established in each time period relative to the resource item quantity established in the previous time period, and according to The 62.2% growth rate of the number of resource items created relative to the number of resource items created in the previous 72 to 96 hours is greater than the first predetermined growth threshold of 50%, and it is judged that the input sequence has timeliness.

For another example, the predetermined condition includes the following two conditions: i) including the sudden semantic feature keywords contained in the input sequence; ii) the number of searches of the input sequence and the keywords contained in it increases with time in any two time periods The rate exceeds 100%; in step S1, the sorting device obtains the input sequence "earthquake gate" containing the sudden characteristic keyword "earthquake", and obtains the input sequence "earthquake gate" and the keyword "earthquake" contained in the input sequence. The change information of the number of searches over time is shown in Table 1 below:

Table 1

Then the sorting device analyzes the input sequence "earthquake gate" and the time-dependent change information of the retrieval times of the keyword "earthquake" contained in the input sequence, and determines that the retrieval times of the input sequence and the keywords contained in it are between any two times. The growth rates over time in the segment are all less than 100%; then the sorting device judges that the input sequence "earthquake gate" has no timeliness.

It should be noted that the above examples are only to better illustrate the technical solution of the present invention, rather than limit the present invention. Those skilled in the art should understand that any method that judges whether the input sequence has timeliness based on whether the timeliness information satisfies the predetermined condition Implementation methods should all be included within the scope of the present invention.

In step S3, when it is judged that the input sequence is time-sensitive, the sorting device performs time-sensitive sorting on the resource items retrieved based on the input sequence, so as to provide the sorted resource items to the user.

Wherein, the sorting device sorts the resource items retrieved based on the input sequence in a time-sensitive manner, so as to provide the sorted resource items to the user, including but not limited to:

1) The sorting device selects a part of resource items from a plurality of resource items retrieved based on the input sequence, and places the selected resource items in a predetermined sorting position, and then sorts unselected resource items; then , the sorting device provides the sorted resource items to the user. Wherein, the sorting device can select the above-mentioned part of resource items according to various factors such as establishment or update time of resource items, click rate of resource items, predetermined quality of resource items, matching degree of resource items and input sequence, etc.

For example, the predetermined sorting position is the top three sorting positions, and the sorting device searches based on the input sequence to obtain multiple resource items, and selects from the multiple resource items the one whose establishment time is shorter than 24 hours and matches the input sequence The resource items whose degree of matching is greater than the predetermined matching degree threshold, then, the sorting device further selects the resource items with the top three click-through rates from the selected resource items, and sorts the resource items with the top three click-through rates according to the click-through rate from high to low The sequence is placed in the top three sorting positions; then, the sorting device uses the existing technology to sort the remaining resource items, and provides the determined top three resource items and other resource items after sorting to the user.

2) The sorting device selects a part of resource items from a plurality of resource items retrieved based on the input sequence, and performs weighting processing on the selected resource items to sort according to the weighted processing results, and The sorted resource items are provided to the user. The manner of sorting through weighting processing will be described in detail with reference to the embodiment shown in FIG. 2 , and will not be repeated here.

It should be noted that the above examples are only to better illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Those skilled in the art should understand that when it is judged that the input sequence is time-sensitive, any The resource items retrieved in sequence are time-sensitively sorted, so as to provide the sorted resource items to the user, all of which should be included in the scope of the present invention.

It should be further explained that, when it is judged that the input sequence has timeliness, the sorting device or other devices perform prior art sorting processing on the resource items retrieved based on the input sequence, so that the sorted resource items provided to the user; or, when it is judged that the input sequence is time-sensitive, the sorting device does not perform any operation.

As one of the preferred manners, the sorting device determines the resource item corresponding to the sorted position of the resource item to be displayed according to the number of resource items to be displayed and the sorted position of the resource item to be displayed. For example, if the user equipment sends an input sequence for querying to the search engine, the sorting device in the search engine can present 10 resource items based on the number of resource items that the user equipment can present at one time, and the resource items obtained based on the input sequence Select to make a selection, and determine the resource items corresponding to the top 10 ranking positions to feed back to the user equipment, so that the user equipment will provide the top 10 resource items to the user; then, when the user equipment sends a page turning request to the search engine, The search engine again selects from the resource items obtained based on the input sequence, and determines the resource items corresponding to ranks 11 to 20 to feed back to the user equipment, so that the user equipment provides the resource items ranked 11 to 20 to user.

As one of the preferred manners, after sorting the resource items obtained based on the input sequence, the sorting device selects some resource items in the sorted positions and provides them to the user according to the sorted position of the resource items presented to the user as required. For example, when the user equipment sends an input sequence for query to the search engine, the sorting device in the search engine sorts the resource items obtained based on the input sequence, and then provides 10 sorted resource items at a time based on the default rule , provide the top 10 resource items to the user device, so that the user device will provide the top 10 resource items to the user; then, when the user device item search engine sends a request to provide resource items on the third page, the search The engine provides the resource items ranked 21st to 30th to the user equipment according to the resource items corresponding to the 3rd page being 21st to 30th resource items, so that the user equipment will sort the resource items of the 21st to 30th resource items provided to the user.

As one of the preferred solutions of the present invention, in the present invention, the following steps S41 (not shown in the figure), step S42 (not shown in the figure) are respectively performed for at least one sudden increase keyword contained in the pre-established user demand sudden increase keyword library shown) and step S43 (not shown).

In step S41 , the sorting device acquires click information of one or more users on the sorted resource items obtained based on the currently processed sudden increase keyword or an input sequence containing the currently processed sudden increase keyword. Preferably, the click information includes the click amount and/or click rate of each sorted resource item obtained based on the currently processed sudden increase keyword or an input sequence containing the currently processed sudden increase keyword.

Specifically, the sorting device obtains click information provided by other devices; or, the sorting device obtains the click information by monitoring and counting the user's click behavior based on the sorted resource items within a predetermined period of time.

Next, in step S42, the sorting device judges whether the timeliness of the sudden increase keyword has expired according to the click information obtained in step S41.

Among them, the way for the sorting device to judge whether the timeliness of the sudden increase keyword has expired includes but is not limited to:

1) The sorting device obtains the judgment evaluation value according to the click-through rate or click-through rate of the resource item assigned to the predetermined sorting position or selected for weighting processing in the aforementioned step S3, combined with the click-through rate or click-through rate of other resource items; And according to whether the judging evaluation value is greater than or smaller than the first predetermined judging threshold, it is judged whether the timeliness of the currently processed sudden increase keyword has expired.

For example, the number of clicks obtained by the sorting device on the seven resource items retrieved based on the sudden increase keyword "May CPI" obtained by the sorting device in step S41 is shown in Table 2 below:

resource item hits resource item A 1230 resource item B 234 resource item C 123 resource item D 467 resource item E 986 resource item F 209 resource item G 78

Table 2

Wherein, resource items A and E are resource items assigned to predetermined sorting positions, and resource item D is a resource item selected for weighting processing, then the sorting device compares the average value of hits of resource items A, D, and E with the resource item The difference 733 of the mean values of clicks of B, C, F and G is used as the judgment evaluation value, and according to the judgment evaluation value being greater than the first predetermined judgment threshold 500, it is judged that the timeliness of the sudden increase keyword currently processed has not expired.

2) The sorting device is based on the click information of the resource items whose establishment or update time is shorter than the first predetermined time threshold in each resource item, combined with the click information of the resource items whose establishment or update time is longer than the second predetermined time threshold among each resource item, to obtain a judgment evaluation value; and according to whether the judgment evaluation value is greater than or smaller than a second predetermined judgment threshold, it is judged whether the timeliness of the currently processed sudden increase keyword has expired.

For example, the sorting device takes the ratio of the click information of the resource items whose establishment or update time is shorter than 3 days and the click information of the resource items whose establishment or update time is longer than 90 days in each resource item as the judgment evaluation value, and according to the judgment evaluation value Whether it is greater than or less than a second predetermined judgment threshold is used to judge whether the timeliness of the currently processed sudden keyword has expired.

It should be noted that the above examples are only to better illustrate the technical solution of the present invention, rather than limit the present invention. Those skilled in the art should understand that any implementation of judging whether the timeliness of a sudden increase keyword has expired based on the click information methods, should be included within the scope of the present invention.

In step S43, when it is judged that the timeliness of the sudden increase keyword has expired, the sorting device deletes the sudden increase keyword from the user demand sudden increase keyword database.

Preferably, when it is judged that the timeliness of the sudden increase keyword has not expired, the sorting device does not perform any operation on the sudden increase keyword.

As one of the preferred solutions of this embodiment, the step S2 further includes step S21 (not shown in the figure) and step S22 (not shown in the figure).

In step S21, the sorting device judges whether the input sequence is time-sensitive according to whether the time-sensitive feature information satisfies a predetermined condition. Wherein, the step of the sorting device judging whether the input sequence has timeliness according to whether the timeliness feature information satisfies the predetermined condition has been described in detail in the aforementioned step S2, and will not be repeated here.

In step S22, when it is judged that the input sequence is time-sensitive, the sorting device adds the input sequence and/or keywords in the input sequence into the user demand sudden increase keyword library. Preferably, the sorting device also adds time-sensitive feature information corresponding to the input sequence into the user demand sudden increase keyword library; more preferably, the sorting device also adds information obtained based on the time-sensitive feature information, for example, the number of resource items or The growth rate of the search times of the input sequence over time, etc., are added to the keyword database for sudden increase in user demand.

The method according to the present invention can identify the input sequence with timeliness, and can perform timeliness sorting on the resource items retrieved based on the input sequence with timeliness, so as to sort the resource items that may be related to the timeliness of the input sequence The sorting is advanced to better meet the user's sudden retrieval needs.

Fig. 2 is a flowchart of a method for sorting resource items according to a preferred embodiment of the present invention. The method according to this embodiment includes step S1, step S2 and step S3. Wherein, step S3 further includes a timeliness sorting step S31.

Step S1 and step S2 have been described in detail with reference to the embodiment shown in FIG. 1 , and are included here by reference, and will not be repeated here.

Next, in step S3, when it is judged that the input sequence is time-sensitive, the sorting device performs time-sensitive sorting on the resource items retrieved based on the input sequence, so as to provide the sorted resource items to the user . In this embodiment, the timeliness sorting step S31 in step S3 includes selecting at least one resource item from among them according to the freshness-related information of each resource item, and performing weighting processing on the selected at least one resource item, so that according to The step of sorting the processing results after the weighting processing.

Wherein, the freshness related information includes related information that can be used to determine whether the establishment or update time of each resource item belongs to the aging period of the input sequence. Preferably, the freshness-related information includes at least one of the following:

1) The creation or update time of the resource item and the user demand surge time of the predetermined input sequence. Wherein, the sorting device can obtain the establishment or update time of each resource item stored in advance, for example, the sorting device or other devices store the establishment or update time of each resource item while grabbing each resource item, for subsequent use; , the establishment or update time may be the establishment or update time of each resource item in its network, or the time when the sorting device or other devices capture and store or update each resource item. Wherein, before step S31, the user demand surge time of the predetermined input sequence can be determined by a sorting device or other devices according to at least one of the following factors: a) The number of resource items obtained by searching based on the input sequence varies with Time change information, the change information of the number of resource items over time has been described in detail in the part of the description of the time-sensitive feature information in the embodiment shown in Figure 1, and will not be repeated here; The change information of the number of resource items over time is used to determine the growth intensity of the number of resource items over time, so as to determine the time of sudden increase in user demand. The starting time point of the time period of the growth threshold is used as the user demand sudden increase time; b) the input sequence and/or the information on the change of the number of searches over time of the keywords contained in the input sequence, and the information on the change of the number of searches over time has been In the embodiment shown in FIG. 1 , the part of the timeliness feature information is described in detail, and will not be repeated here; the sorting device or other devices determine the number of times retrieved over time according to the change information of the number of times retrieved over time. The growth intensity of the user to determine the time of sudden increase in user demand.

Specifically, the sorting device selects the resource items whose establishment or update time is later than the user demand surge time from among the resource items retrieved based on the input sequence, and performs weighting processing on the selected resource items, so that according to the weighting processing The final processing results are sorted.

For example, if the predetermined user demand surge time is May 12, the sorting device selects the resource items whose creation time is later than May 12 from among the resource items obtained by searching based on the input The weighted value is used to carry out weighting processing on each selected resource item, so as to sort according to the processing result after the weighting processing.

2) The resource items included in the historical retrieval records of the input sequence.

Specifically, the sorting device selects resource items that are not included in the historical retrieval records of the input sequence from the resource items obtained by searching based on the input sequence, and performs weighting processing on the selected resource items, so that according to the weighted processing The processing results are sorted. Since the selected resource item is not included in the historical retrieval record of the data sequence, it can be considered that the creation time of the selected resource item is later than the last update time of the historical retrieval record, so the resource item selected based on the historical retrieval record The creation or update time should be within the aging period of the input sequence.

It should be noted that the above examples are only to better illustrate the technical solutions of the present invention, rather than limiting the present invention. Those skilled in the art should understand that any information related to the freshness of each resource item is selected from which at least One resource item, and performing weighting processing on at least one selected resource item, so as to sort according to the processing result after the weighting processing, should be included in the scope of the present invention.

It should be further explained that the sorting device can first select resource items that require all weighted processing and perform weighted processing, and then sort all resource items that have undergone weighted processing and that have not undergone weighted processing; The weighted resource item is to perform weighted processing on the resource item and insert the resource item into the sorted resource item.

As one of the preferred solutions of this embodiment, the method of this embodiment further includes steps S5 and S6, and the weighting processing step in step S31 further includes weighting the selected resource items that need to be weighted according to the weighting value processing steps.

In step S5 executed after step S2, the sorting device determines the change intensity of the user demand of the input sequence according to the aging information obtained in step S1.

Specifically, the way for the sorting device to determine the intensity of change in user demand of the input sequence includes but is not limited to:

1) When the timeliness information includes the query result obtained by the sorting device in the user demand sudden increase keyword database, and the query result includes the user demand change intensity of the sudden increase keyword obtained from the query, the sorting device according to the query result The user demand change strength of the sudden increase keyword is used to determine the user demand change strength of the input sequence.

For example, if the input sequence is "how about the Fukushima earthquake", in step S1, the query result obtained by the sorting device contains the sudden increase keyword "Fukushima earthquake" and the corresponding user demand change intensity is level 3, then the sorting device directly The user demand change intensity corresponding to the sudden increase keyword "Fukushima earthquake" is used as the user demand change intensity of the input sequence "how about the Fukushima earthquake".

For another example, if the input sequence is "Fukushima Earthquake Nuclear Power Plant", in step S1, the query result obtained by the sorting device includes the sudden increase keyword "Fukushima Earthquake" corresponding to a user demand change intensity of 4.76, and the sudden increase keyword "Nuclear Power Plant The user demand change intensity corresponding to "is 2.35, then the sorting device takes the sum of user demand change intensity corresponding to the sudden increase keyword as 7.11 as the user demand change intensity of the input sequence "Fukushima Earthquake Nuclear Power Plant".

2) The timeliness information includes timeliness feature information, and the sorting device determines the change intensity of the user demand of the input sequence according to the timeliness feature information.

For example, the time-sensitive feature information includes information on the change of the number of resource items over time obtained by searching based on the input sequence, and the sorting device determines the input The intensity of user demand change for the sequence. Wherein, the machine learning model for determining the change intensity of the user demand can be obtained by training the machine learning model with a large number of input sequences that have determined the change intensity of the user demand and the information on the change of the number of resource items over time.

For another example, in step S1, the time-sensitive feature information obtained by the sorting device based on the input sequence "Fukushima Earthquake" includes the following two items: 1) contains the burst feature keyword "earthquake" and the burst feature semantic structure "proper noun + 2) the change information of the retrieved times of the input sequence "Fukushima Earthquake" over time, the change information is shown in Table 3 below:

point in time Searched times March 11 at 11:00 324 March 11 at 12 o'clock 235 March 11 at 13:00 375 March 11 at 14:00 2045 March 11 at 15:00 15247 March 11 at 16:00 16874 March 11 at 17:00 17896

table 3

Then the sorting device calculates the difference value of the number of searches at two adjacent time points one by one, and obtains the maximum difference = 15247-2045 = 13202, and then, the sorting device obtains the ratio of the maximum difference to the maximum number of searches = 13202/17896 =73.77%, then, the sorting device determines that 73.77% belongs to the predetermined range of 70-80%, and obtains the user demand change intensity corresponding to the predetermined range of 70-80% as the second level. Moreover, the sorting device determines that the burst feature information contains burst feature keywords and the burst feature semantic structure according to the predetermined rule that the level adjustment method corresponding to the burst semantic feature information is to raise one level. The level adjustment rule corresponding to the burst semantic feature information of the keyword "earthquake" and the burst feature semantic structure "proper noun + common noun" is one level up. Then the sorting device adjusts the grade obtained based on the change information of the number of retrieved times over time according to the grade adjustment rule determined based on the burst semantic feature information, and finally determines that the user demand change intensity of the input sequence "Fukushima earthquake" is the first grade.

Preferably, the sorting device adds the input sequence and its user demand change intensity to the user demand sudden increase keyword library.

It should be noted that the above examples are only to better illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Those skilled in the art should understand that any implementation method for determining the intensity of changes in user needs of an input sequence based on time-sensitive information, All should be included within the scope of the present invention.

Next, in step S6, the sorting device determines a weighted value according to the change intensity of the user demand.

For example, when the change intensity of the user's demand is a numerical value, the numerical value is directly used as the weighted value, or the numerical value is multiplied by a predetermined adjustment factor, and the product is used as the weighted value, etc.

For another example, when the change intensity of the user demand is at a level, a predetermined weighted value corresponding to the level is obtained.

It should be noted that the above examples are only to better illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Those skilled in the art should understand that any implementation of determining the weighted value according to the change intensity of the user's needs should be included within the scope of the present invention.

Next, the sorting device executes step S3 including step S31. In this embodiment, the weighting processing step in step S31 further includes performing weighting processing on the selected resource items that need to be weighted according to the weighting value determined in step S6. step.

Preferably, the weighting processing step in step S31 further includes the step of performing weighting processing on the selected resource items according to the weight value determined in step S6 and in combination with the establishment or update time of each resource item in the selected resource items .

Specifically, the sorting device adjusts the weight value according to the establishment or update time of each resource item, and performs weight processing on the selected resource item according to the adjusted weight value.

For example, the weighted value obtained by the sorting device in step S6 is 2.3, and the establishment time of a resource item that needs to be weighted is obtained by the sorting device at 13 o'clock on May 24th, then the sorting device will calculate according to the establishment time and the current time in May If the time interval at 20:00 on the 24th is less than the predetermined time interval of 1 day, the weighted value is increased by 0.1 to obtain an adjusted weighted value of 2.4, and the resource item is weighted according to the adjusted weighted value of 2.4.

For another example, the weighted value obtained by the sorting device in step S6 is 0.1, and the establishment time of a resource item that needs to be weighted is obtained by the sorting device as May 24. Then, the sorting device determines that the establishment time corresponds to the input sequence The time interval between the user demand surge time May 22 is 2 days, and the sorting device uses the product of the weighted value and the time interval of 0.2 as the adjusted weighted value to perform weighting processing on the resource item.

It should be noted that the above examples are only to better illustrate the technical solution of the present invention, rather than limit the present invention. Those skilled in the art should understand that any combination of resource items selected according to weighted values The establishment or update time of the resource item, and the implementation manner of weighting the selected resource items should all be included in the scope of the present invention.

According to the method of this embodiment, timeliness sorting is further performed according to the freshness of resource items, so as to obtain a better timeliness sorting result.

Fig. 3 is a schematic structural diagram of a sorting device for sorting resource items according to one aspect of the present invention. The ranking device according to the present invention includes a timeliness information acquisition device 1 , a first judging device 2 and a timeliness sorting device 3 .

The aging information acquiring device 1 acquires aging information related to the input sequence according to the input sequence from the user.

Specifically, the timeliness information acquiring device 1 acquires an input sequence provided by other devices or equipment, or acquires an input sequence from a user in real time, and acquires timeliness information related to the input sequence according to the input sequence. Wherein, the methods for obtaining the timeliness information related to the input sequence by the timeliness information acquisition device 1 include but are not limited to:

1) The timeliness information acquisition device 1 searches the pre-established user demand sudden increase keyword database according to the input sequence, and takes the query result obtained from the query as the timeliness information. Wherein, the query result includes any of the following items: a) It is used to indicate that the input sequence or the keywords contained in the input sequence can be obtained by querying the user demand sudden keyword database, for example, "Y", " 1" etc.; b) It is used to indicate that the input sequence or the identification information of the keywords contained in the input sequence cannot be obtained from the user demand sudden keyword database query, for example, "N", "0" and so on. Preferably, the query result also includes at least one of the following: a) the sudden increase keyword in the user demand sudden increase keyword library obtained by querying the user demand sudden increase keyword database; Keyword-related information, for example, input sequence or user demand change intensity of keywords, etc.

2) The timeliness information acquisition device 1 includes a first sub-acquisition device (not shown in the figure), and the first sub-acquisition device acquires timeliness feature information related to the input sequence according to the input sequence as the timeliness information. Wherein, the time-sensitive feature information includes feature information that can be used to correlate with the input sequence and can be used to judge whether the input sequence is a time-sensitive input sequence. Preferably, the time-sensitive feature information includes but not limited to at least the following item:

a) information on the change in the number of resource items over time obtained by searching based on the input sequence; for example, the information on the change in the number of resource items over time includes the following information: the resource items obtained by searching based on the input sequence contain 50 Resource items created within the previous 96 to 120 hours, 48 resource items created within the previous 72 to 96 hours, 127 resource items created within the previous 48 hours to the previous 72 hours, 309 resource items created within the previous 24 to Resource items created within the previous 48 hours and 1045 resource items created within 24 hours; as another example, the information on the change in the number of resource items over time includes the relative quantity of resource information created within each predetermined time period. The change value of the amount of resource information established in the adjacent time period, etc. Preferably, the first sub-acquisition means obtains information on changes in the number of resource items over time by statistically analyzing the creation or update time of the resource items retrieved based on the input sequence.

b) information on the number of times the input sequence and/or keywords included in the input sequence are retrieved over time; for example, the input sequence or the number of times the keywords included in the input sequence are retrieved within each predetermined time period or at a predetermined time point ; Another example is the difference between each predetermined time period or predetermined time point relative to its adjacent time period or time point's input sequence or the number of times keywords are retrieved. Preferably, the first sub-acquisition means searches the local storage device based on the input sequence or the keywords contained in the input sequence, so as to obtain the pre-stored data of the times of retrieval corresponding to the input sequence or the keywords contained in the input sequence over time. Change information; or, the first sub-acquisition means sends a request to other devices or equipment to obtain information on the number of searches over time corresponding to the input sequence or the keywords contained in the input sequence; or, the first sub-acquisition means through the Monitoring and counting the number of retrievals of the input sequence or keywords included in the input sequence within a predetermined period of time to obtain information on changes in the retrieval frequency over time.

c) Semantic burst feature information of the input sequence obtained based on semantic analysis of the input sequence. Wherein, the burst semantic feature information includes burst feature keywords and/or burst feature semantic structure. For example, the burst semantic feature information includes time keywords such as "latest, today, this week, recent, certain year, month, and day", or includes time keywords such as "earthquake, car accident, death, fire, hacking incident, suicide door" and other emergency keywords; for another example, the emergency semantic feature structure includes structures such as proper nouns + common nouns; for another example, the emergency semantic feature information includes such as "Fukushima earthquake" and other emergency feature keywords and Semantic feature information of burst feature semantic structure. Preferably, the first sub-acquisition means searches the local storage device based on the input sequence to obtain pre-stored burst semantic feature information corresponding to the input sequence; or, the first sub-acquisition means sends a request to other means or equipment to obtain burst semantic feature information corresponding to the input sequence; or, the first sub-acquisition means performs semantic analysis on the input sequence to obtain burst semantic feature information of the input sequence.

3) The aging information acquisition device 1 includes a query device (not shown in the figure), a definition device (not shown in the figure) and a second sub-acquisition device (not shown in the figure). The query device queries the sudden increase keywords related to the input sequence in the pre-established user demand sudden increase keyword library according to the input sequence; when the sudden increase keywords related to the input sequence are found, the definition device will query The result is used as the timeliness information; when the sudden increase keyword related to the input sequence cannot be found, the first sub-acquisition means acquires the timeliness feature information related to the input sequence according to the input sequence, as the Steps for aging information. Wherein, the manner in which the first sub-acquisition means acquires the time-sensitive feature information related to the input sequence according to the input sequence is the same as or similar to the way in which the first sub-acquisition means acquires the time-sensitive feature information in 2) above, and will not be repeated here. .

It should be noted that the above examples are only to better illustrate the technical solution of the present invention, rather than limit the present invention. Those skilled in the art should understand that any time-sensitive information related to the input sequence obtained according to the input sequence from the user All implementation methods should be included within the scope of the present invention.

Next, the first judging means 2 judges whether the input sequence has timeliness according to whether the timeliness information satisfies a predetermined condition. Wherein, the predetermined conditions include the conditions that need to be satisfied by the timeliness information obtained based on the input sequence that can be determined to be timeliness; When the query results are obtained, the predetermined conditions include identification information such as "Y" or "1" in the query results, and when the timeliness information includes timeliness feature information, the predetermined conditions include conditions that the timeliness feature information needs to meet, etc. Preferably, based on the input The faster the number of resource items retrieved by the sequence increases with time, the faster the number of retrievals of the input sequence or keywords contained in the input sequence increases with time, and the stronger the burst semantic feature contained in the burst semantic feature information , the sorting device is more likely to judge the input sequence as a time-sensitive input sequence.

For example, the predetermined condition includes that the query result contains the identification information "Y" or "1", and the timeliness information obtained by the aging information acquisition device 1 based on the input sequence includes the query result containing the identification information "Y"; then the first judging device 2 according to the The timeliness information satisfies the predetermined condition, and it is judged that the input sequence has timeliness.

For another example, the predetermined condition includes that the growth rate of the number of resource items established in any time period relative to the number of resource items established in the previous time period exceeds 50%; The change information of the number of resource items over time includes the following information: among the resource items retrieved based on the input sequence, there are 50 resource items established within the previous 96 to 120 hours, 48 resource items established within the previous 72 to 96 hours 127 resource items created within the previous 48 hours to the previous 72 hours, 309 resource items created within the previous 24 to 48 hours, and 1045 resource items created within 24 hours; then A judging device 2 analyzes the change information of the quantity of resource items over time, determines the growth rate of the quantity of resource items established in each time period relative to the quantity of resource items established in the previous time period, and according to the number of resource items established in the previous 48 hours to the previous 72 hours The 62.2% growth rate of the number of resource items created within the previous 72 to 96 hours relative to the number of resource items created within the previous 96 hours is greater than the first predetermined growth threshold of 50%, and it is judged that the input sequence has timeliness.

For another example, the predetermined conditions include the following two conditions: i) the input sequence contains sudden semantic feature keywords; ii) the number of searches of the input sequence and the keywords contained in it is the growth rate over time in any two time periods Both exceeded 100%. The time-sensitive information acquisition device obtains the input sequence "earthquake gate" containing the sudden characteristic keyword "earthquake", and obtains the input sequence "earthquake gate" and the information on the number of retrievals of the keyword "earthquake" contained in the input sequence over time as follows Table 4 shows:

Table 4

Then the first judging device 2 analyzes the input sequence "earthquake gate" and the time-dependent change information of the retrieved times of the keyword "earthquake" contained in the input sequence, and determines that the input sequence and the retrieved times of the keywords contained in it are within any The growth rates over time in the two time periods are both less than 100%; then the first judging device 2 judges that the input sequence "earthquake gate" has no timeliness.

It should be noted that the above examples are only to better illustrate the technical solution of the present invention, rather than limit the present invention. Those skilled in the art should understand that any method that judges whether the input sequence has timeliness based on whether the timeliness information satisfies the predetermined condition Implementation methods should all be included within the scope of the present invention.

When it is judged that the input sequence has timeliness, the timeliness sorting device 3 performs timeliness sorting on the resource items retrieved based on the input sequence, so as to provide the sorted resource items to the user.

Wherein, the timeliness sorting device 3 performs timeliness sorting on the resource items retrieved based on the input sequence, and the ways of providing the sorted resource items to the user include but not limited to:

1) The timeliness sorting device 3 selects a part of resource items from a plurality of resource items retrieved based on the input sequence, and places the selected resource items in a predetermined sorting position, and then sorts the unselected resource items ; Next, the sorting device provides the sorted resource items to the user. Among them, the timeliness sorting device 3 can select the above-mentioned part of resource items according to various factors such as establishment or update time of resource items, click rate of resource items, predetermined quality of resource items, matching degree of resource items and input sequence, etc.

For example, the predetermined sorting position is the first three sorting positions, and the sorting device searches based on the input sequence to obtain multiple resource items, and the timeliness sorting device 3 selects from the multiple resource items that the construction time is shorter than 24 hours and is consistent with the input The resource items whose matching degree of the sequence is greater than the predetermined matching degree threshold, then, the timeliness sorting device 3 further selects the resource items with the top three click rates from the selected resource items, and sorts the resource items with the top three click rates according to the click rate. Rates are placed in the top three sorting positions in order from high to low; then, the aging sorting device 3 uses the prior art to sort the remaining resource items; then, the sorting device sorts the determined top three resource items and the sorted Other resource items are provided to the user.

2) The timeliness sorting device 3 selects a part of resource items from a plurality of resource items retrieved based on the input sequence, and performs weighting processing on the selected resource items, so as to sort according to the weighted processing results, And provide the sorted resource items to the user. The manner of sorting through weighting processing will be described in detail with reference to the embodiment shown in FIG. 2 , and will not be repeated here.

It should be noted that the above examples are only to better illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Those skilled in the art should understand that when it is judged that the input sequence is time-sensitive, any The resource items retrieved in sequence are time-sensitively sorted, so as to provide the sorted resource items to the user, all of which should be included in the scope of the present invention.

It should be further explained that, when it is judged that the input sequence has timeliness, the sorting device or other devices perform prior art sorting processing on the resource items retrieved based on the input sequence, so that the sorted resource items provided to the user; or, when it is judged that the input sequence is time-sensitive, the sorting device does not perform any operation.

As one of the preferred ways, the timeliness sorting device 3 determines the resource item corresponding to the resource item sorting position that needs to be presented according to the number of resource items that need to be presented and the sorting position of the resource items that need to be presented. For example, if the user equipment sends an input sequence for querying to the search engine, then the timeliness sorting device 3 in the search engine can present 10 resource items at one time according to the user equipment, and the resource items obtained based on the input sequence Item selection, and determine the resource items corresponding to the top 10 ranking positions to feed back to the user equipment, so that the user equipment will provide the top 10 resource items to the user; then, when the user equipment sends the search engine to turn the page request, the timeliness sorting device 3 selects again from the resource items obtained based on the input sequence, and determines the resource items corresponding to the ranks 11 to 20 to feed back to the user equipment, so that the user equipment ranks the resource items of the ranks 11 to 20 Resource items are offered to users.

As one of the preferred ways, after sorting the resource items obtained based on the input sequence, the timeliness sorting device 3 selects some resource items in the sorted position and provides them to the user according to the sorting positions of the resource items presented to the user as required. For example, if the user equipment sends an input sequence for querying to the search engine, the timeliness sorting device 3 in the search engine sorts the resource items obtained based on the input sequence, and provides 10 sorted resource items at a time based on the default According to the rule, provide the top 10 resource items to the user equipment, so that the user equipment will provide the top 10 resource items to the user; then, when the user equipment item search engine sends a request to provide resource items on the third page, Then the timeliness sorting device 3 provides the resource items ranked 21st to 30th to the user equipment according to the resource items corresponding to the 3rd page being 21st to 30th resource items, so that the user equipment will sort the 21st to 30th resource items The item's resource item is provided to the user.

As one of the preferred solutions of the present invention, the sorting device according to the present invention also includes an update device (not shown in the figure), which updates according to the click information acquisition device (not shown in the figure) and the second judging device (not shown in the figure) included in the update device. ) and a deleting device (not shown in the figure), to respectively perform operations on at least one sudden increase keyword contained in the user demand sudden increase keyword library.

The click information acquiring means acquires click information of one or more users on sorted resource items obtained based on the currently processed sudden increase keyword or an input sequence containing the currently processed sudden increase keyword. Preferably, the click information includes the click amount and/or click rate of each sorted resource item obtained based on the currently processed sudden increase keyword or an input sequence containing the currently processed sudden increase keyword.

Specifically, the click information acquisition means acquires click information provided by other devices; or, the click information acquisition means monitors and counts click behaviors of users based on the sorted resource items within a predetermined period of time to obtain click information.

Next, the second judging means judges, according to the click information obtained in the aforementioned step S41, whether the timeliness of the sudden increase keyword has expired.

Wherein, the way for the second judging device to judge whether the timeliness of the sudden increase keyword has expired includes but is not limited to:

1) The second judging means obtains the judging evaluation value according to the click-through rate or click-through rate of the resource item assigned to the predetermined sorting position or selected to be weighted by the aging sorting means 3, combined with the click-through rate or click-through rate of other resource items ; and according to whether the judgment evaluation value is greater than or less than the first predetermined judgment threshold, it is judged whether the timeliness of the sudden increase keyword currently processed has expired.

For example, the amount of clicks obtained by the user on the seven resource items retrieved based on the sudden increase keyword "CPI in May" obtained by clicking on the information acquisition device is shown in Table 5 below:

resource item hits resource item A 1230 resource item B 234 resource item C 123 resource item D 467 resource item E 986 resource item F 209 resource item G 78

table 5

Wherein, resource items A and E are resource items assigned to a predetermined sorting position, and resource item D is a resource item selected for weighting processing, then the second judging means compares the average value of hits of resource items A, D, and E with The difference 733 of the average value of hits of resource items B, C, F, and G is used as the judgment evaluation value, and according to the judgment evaluation value being greater than the first predetermined judgment threshold 500, it is judged that the timeliness of the sudden increase keyword currently processed has not expired.

2) The second judging means is based on the click information of resource items whose establishment or update time is shorter than the first predetermined time threshold in each resource item, combined with the click information of resource items whose establishment or update time is longer than the second predetermined time threshold in each resource item information to obtain a judgment evaluation value; and according to whether the judgment evaluation value is greater than or less than a second predetermined judgment threshold, it is judged whether the timeliness of the currently processed sudden increase keyword has expired.

For example, the second judging means takes the ratio of the click information of the resource items whose establishment or update time is shorter than 3 days and the click information of the resource items whose establishment or update time is longer than 90 days in each resource item as the judgment evaluation value, and according to the judgment Whether the evaluation value is greater than or less than a second predetermined judgment threshold is used to judge whether the timeliness of the currently processed sudden keyword has expired.

It should be noted that the above examples are only to better illustrate the technical solution of the present invention, rather than limit the present invention. Those skilled in the art should understand that any implementation of judging whether the timeliness of a sudden increase keyword has expired based on the click information methods, should be included within the scope of the present invention.

When it is judged that the timeliness of the sudden increase keyword has expired, the deleting device deletes the sudden increase keyword in the user demand sudden increase keyword database.

Preferably, when it is judged that the timeliness of the sudden increase keyword has not expired, the sorting device does not perform any operation on the sudden increase keyword.

As one of the preferred solutions of this embodiment, the first judging device 2 further includes a sub-judging device (not shown in the figure) and an adding device (not shown in the figure).

The sub-judgment means judges whether the input sequence is time-sensitive according to whether the time-sensitive characteristic information satisfies a predetermined condition. Wherein, the manner in which the sub-judgment means judges whether the input sequence is time-sensitive according to whether the time-sensitive feature information satisfies the predetermined condition is the same or similar to the way that the first judging means 2 judges whether the input sequence is time-sensitive according to whether the time-sensitive feature information satisfies the predetermined condition , which will not be repeated here.

When judging that the input sequence has timeliness, the adding device adds the input sequence and/or keywords in the input sequence to the user demand sudden increase keyword library. Preferably, the adding means also adds the time-sensitive feature information corresponding to the input sequence into the user demand sudden increase keyword library; more preferably, the adding means also adds information obtained based on the time-sensitive feature information, for example, the number of resource items or The growth rate of the search times of the input sequence over time, etc., are added to the keyword database for sudden increase in user demand.

The sorting device according to the present invention can identify the input sequence with timeliness, and can perform timeliness sorting on the resource items retrieved based on the input sequence with timeliness, so as to sort resources that may be related to the timeliness of the input sequence Items are sorted in advance, which better meets the user's sudden retrieval needs.

Fig. 4 is a schematic structural diagram of a sorting device for sorting resource items according to a preferred embodiment of the present invention. The method according to this embodiment includes an aging information acquiring means 1 , a first judging means 2 and an aging sorting means 3 . Wherein, the aging sorting device 3 further includes a sub-sorting device 31 .

The aging information acquiring device 1 and the first judging device 2 have been described in detail with reference to the embodiment shown in FIG. 2 , and are included here by reference, and will not be repeated here.

When it is judged that the input sequence has timeliness, the timeliness sorting device 3 performs timeliness sorting on the resource items retrieved based on the input sequence, so as to provide the sorted resource items to the user. In this embodiment, the sub-sorting device 31 in the timeliness sorting device 3 performs the operation of the timeliness sorting, specifically, the sub-sorting device 31 selects at least one resource item from among them according to the freshness related information of each resource item , and performing weighting processing on at least one selected resource item, so as to perform a step of sorting according to a processing result after the weighting processing.

Wherein, the freshness related information includes related information that can be used to determine whether the establishment or update time of each resource item belongs to the aging period of the input sequence. Preferably, the freshness-related information includes at least one of the following:

1) The creation or update time of the resource item and the user demand surge time of the predetermined input sequence. Wherein, the sorting device can obtain the establishment or update time of each resource item stored in advance, for example, the sorting device or other devices store the establishment or update time of each resource item while grabbing each resource item, for subsequent use; , the establishment or update time may be the establishment or update time of each resource item in its network, or the time when the sorting device or other devices capture and store or update each resource item. Wherein, before the sub-sorting device 31 executes the operation, the user demand surge time of the predetermined input sequence can be determined by the sorting device or other devices according to at least one of the following factors: a) the time obtained by searching based on the input sequence The change information of the number of resource items over time, the information about the change of the number of resource items over time has been described in detail in the part of the description of the time-sensitive feature information in the embodiment shown in Figure 3, and will not be repeated here; the sorting device or other The device determines the growth intensity of the number of resource items over time according to the change information of the number of resource items over time, so as to determine the time of sudden increase in user demand. For example, the sorting device or other devices determine the growth intensity of the number of resource items The starting time point of the time period exceeding the second predetermined growth threshold is taken as the user demand sudden increase time; b) information on the number of searches over time of the input sequence and/or keywords contained in the input sequence, the number of searches over time The change information has been described in detail in the part of the description of the time-sensitive feature information in the embodiment shown in FIG. 3 , and will not be repeated here; The growth intensity of retrieval times over time to determine the time when user demand surges.

Specifically, the sub-sorting device 31 selects the resource items whose establishment or update time is later than the user demand surge time from among the resource items retrieved based on the input sequence, and performs weighting processing on the selected resource items to obtain according to the The processing results after weighting processing are sorted.

For example, if the predetermined user demand surge time is May 12, the sub-sorting device 31 selects the resource items whose creation time is later than May 12 from among the resource items obtained by searching based on the input sequence, and according to The selected resource items are weighted according to a predetermined weight value, so as to be sorted according to the weighted processing results.

2) The resource items included in the historical retrieval records of the input sequence.

Specifically, the sub-sorting device 31 selects resource items that are not included in the historical retrieval records of the input sequence from the resource items retrieved based on the input sequence, and performs weighting processing on the selected resource items, so that according to the weighted The processed results are sorted. Since the selected resource item is not included in the historical retrieval record of the data sequence, it can be considered that the creation time of the selected resource item is later than the last update time of the historical retrieval record, so the resource item selected based on the historical retrieval record The creation or update time should be within the aging period of the input sequence.

It should be noted that the above examples are only to better illustrate the technical solutions of the present invention, rather than limiting the present invention. Those skilled in the art should understand that any information related to the freshness of each resource item is selected from which at least One resource item, and performing weighting processing on at least one selected resource item, so as to sort according to the processing result after the weighting processing, should be included in the scope of the present invention.

It should be further explained that the sub-sorting device 31 may first select resource items that require all weighted processing and perform weighting processing, and then sort all resource items that have undergone weighted processing and have not undergone weighted processing; or, the sub-sorting device 31 Each time a resource item that needs to be weighted is obtained, the resource item is weighted, and the resource item is inserted into the sorted resource items.

As one of the preferred schemes of this embodiment, the sorting device of this embodiment also includes a strength determining device and a weighted value determining device, and the aforementioned sub-sorting device 31 further includes a weighting processing device (not shown) that performs a weighting processing operation, the weighting A step in which the processing device performs weighting processing on the selected resource items that need to be weighted according to the weighting value.

The intensity determination means determines the change intensity of the user demand of the input sequence according to the aging information obtained by the aging information acquiring means 1 .

Specifically, the means for determining the strength of the user demand change of the input sequence include but not limited to:

1) When the timeliness information includes the query result obtained by the sorting device in the user demand sudden increase keyword database, and the query result includes the user demand change intensity of the query sudden increase keyword, the strength determination device according to the query result The user demand change strength of the sudden increase keyword is used to determine the user demand change strength of the input sequence.

For example, if the input sequence is "how about the Fukushima earthquake", and the query result obtained by the time-sensitive information acquisition device 1 contains the sudden increase keyword "Fukushima earthquake", the user demand change intensity corresponding to level 3 is 3, then the intensity determination device will directly The user demand change intensity corresponding to the sudden increase keyword "Fukushima earthquake" is used as the user demand change intensity of the input sequence "how about the Fukushima earthquake".

For another example, the input sequence is "Fukushima Earthquake Nuclear Power Plant", and the query result obtained by the time-sensitive information acquisition device 1 includes the sudden increase keyword "Fukushima Earthquake", the user demand change intensity is 4.76, and the sudden increase keyword "Nuclear Power Plant" corresponds to If the user demand change intensity is 2.35, the intensity determination device takes the sum of user demand change intensity corresponding to the sudden increase keyword as 7.11 as the user demand change intensity of the input sequence "Fukushima Earthquake Nuclear Power Plant".

2) The timeliness information includes timeliness feature information, and the intensity determining means determines the change intensity of the user demand of the input sequence according to the timeliness feature information.

For example, the time-sensitive feature information includes information on the change of the number of resource items over time obtained by searching based on the input sequence, and the strength determination device determines the The intensity of user demand change for the input sequence. Wherein, the machine learning model for determining the change intensity of the user demand can be obtained by training the machine learning model with a large number of input sequences that have determined the change intensity of the user demand and the information on the change of the number of resource items over time.

For another example, the time-sensitive feature information obtained by the time-sensitive information acquisition device 1 based on the input sequence "Fukushima earthquake" includes the following two items: 1) contains the burst feature keyword "earthquake" and the burst feature semantic structure "proper noun + common noun" 2) The change information of the retrieved times of the input sequence "Fukushima Earthquake" over time, the change information is shown in Table 6 below:

point in time Searched times March 11 at 11:00 324 March 11 at 12 o'clock 235 March 11 at 13:00 375 March 11 at 14:00 2045 March 11 at 15:00 15247 March 11 at 16:00 16874 March 11 at 17:00 17896

Table 6

Then the strength determination device calculates the difference value of the number of times retrieved at two adjacent time points one by one, and obtains the maximum difference = 15247-2045 = 13202, and then, the strength determination device obtains the ratio of the maximum difference to the maximum number of times retrieved = 13202 /17896=73.77%, then, the intensity determining means determines that 73.77% belongs to the predetermined range of 70-80%, and obtains the user demand change intensity corresponding to the predetermined range of 70-80% as the second level. Moreover, the intensity determination device determines that the burst semantic feature information contains burst feature keywords and burst feature semantic structure according to the predetermined rule that the level adjustment mode corresponding to the burst semantic feature information is to increase one level. The level adjustment rule corresponding to the burst semantic feature information of the feature keyword "earthquake" and the burst feature semantic structure "proper noun + common noun" is to raise one level. Then the intensity determination device adjusts the level obtained based on the change information of the number of retrieved times over time according to the level adjustment rule determined based on the burst semantic feature information, and finally determines that the user demand change intensity of the input sequence "Fukushima earthquake" is the first level .

Preferably, the sorting device adds the input sequence and its user demand change intensity into the user demand sudden increase keyword library.

It should be noted that the above examples are only to better illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Those skilled in the art should understand that any implementation method for determining the intensity of changes in user needs of an input sequence based on time-sensitive information, All should be included within the scope of the present invention.

Next, the weighted value determining means determines the weighted value according to the change intensity of the user's demand.

For example, when the change intensity of the user's demand is a numerical value, the numerical value is directly used as the weighted value, or the numerical value is multiplied by a predetermined adjustment factor, and the product is used as the weighted value, etc.

For another example, when the change intensity of the user demand is at a level, a predetermined weighted value corresponding to the level is obtained.

It should be noted that the above examples are only to better illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Those skilled in the art should understand that any implementation of determining the weighted value according to the change intensity of the user's needs should be included within the scope of the present invention.

Next, the aging sorting device 3 performs an operation. In this embodiment, the weighting processing device included in the sub-sorting device 31 performs weighting processing on the selected resource items that need to be weighted according to the weight value determined by the weighting determination device. .

Preferably, the weighting processing device further includes a sub-weighting processing device, and the sub-weighting processing device calculates the selected resource item according to the weight value determined by the weight determining device and in combination with the establishment or update time of each resource item in the selected resource item. Steps to carry out the weighting process.

Specifically, the sorting device adjusts the weight value according to the establishment or update time of each resource item, and performs weight processing on the selected resource item according to the adjusted weight value.

For example, if the weight value obtained by the weight determination device is 2.3, and the establishment time of the resource item that needs to be weighted is 13 o'clock on May 24, then the sub-weight processing device will calculate the value based on the establishment time and the current time at 20 o'clock on May 24. If the time interval is less than 1 day, the weighted value is increased by 0.1 to obtain an adjusted weighted value of 2.4, and the resource item is weighted according to the adjusted weighted value of 2.4.

For another example, the weighted value obtained by the weight determination device is 0.1, and the establishment time of the resource item that needs to be weighted is May 24, and then, the sub-weight processing device determines that the establishment time corresponds to the user demand sudden increase time of the input sequence If the time interval of May 22 is 2 days, the sub-weighting processing device uses 0.2, the product of the weight value and the time interval, as the adjusted weight value to perform weighting processing on the resource item.

It should be noted that the above examples are only to better illustrate the technical solution of the present invention, rather than limit the present invention. Those skilled in the art should understand that any combination of resource items selected according to weighted values The establishment or update time of the resource item, and the implementation manner of weighting the selected resource items should all be included in the scope of the present invention.

According to the sorting apparatus of this embodiment, time-sensitive sorting is further performed according to the freshness of resource items, so as to obtain better time-sensitive sorting results.

It will be apparent to those skilled in the art that the invention is not limited to the details of the above-described exemplary embodiments, but that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Accordingly, the embodiments should be regarded in all points of view as exemplary and not restrictive, the scope of the invention being defined by the appended claims rather than the foregoing description, and it is therefore intended that the scope of the invention be defined by the appended claims rather than by the foregoing description. All changes within the meaning and range of equivalents of the elements are embraced in the present invention. Any reference sign in a claim should not be construed as limiting the claim concerned. In addition, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or devices stated in the system claims may also be realized by one unit or device through software or hardware. The words first, second, etc. are used to denote names without implying any particular order.

Claims (21)

1. A computer-implemented method for sorting resource items, wherein the method comprises the steps of: a Acquiring time-sensitive information related to the input sequence according to the input sequence from the user; b. Judging whether the input sequence has timeliness according to whether the timeliness information satisfies a predetermined condition; Wherein, the method also includes the following steps: x When judging that the input sequence has timeliness, perform timeliness sorting on the resource items retrieved based on the input sequence, so as to provide the sorted resource items to the user; Wherein, the timeliness sorting step in the x includes: - Select at least one resource item according to the freshness-related information of each resource item, and perform weighting processing on the selected at least one resource item, so as to sort according to the weighted processing result, wherein the freshness The degree related information includes related information that can be used to determine whether the creation or update time of each resource item belongs to the aging period of the input sequence. 2. The method according to claim 1, wherein said a comprises the steps of: - Acquiring, according to the input sequence, time-sensitive characteristic information related to the input sequence as the time-sensitive information. 3. The method according to claim 1, wherein said a comprises the steps of: - query keywords related to the input sequence in the pre-established user demand sudden increase keyword library according to the input sequence; - When a keyword related to the input sequence is found, the query result is used as the timeliness information. 4. The method according to claim 3, wherein said a further comprises the steps of: - when the keyword related to the input sequence cannot be found, according to the input sequence, time-sensitive characteristic information related to the input sequence is acquired as the time-sensitive information. 5. The method according to claim 4, wherein the aging feature information includes at least one of the following: - information on changes in the quantity of resource items over time obtained by searching based on the input sequence; - information on the number of searches of the input sequence and/or keywords included in the input sequence over time; - burst semantic feature information of the input sequence obtained based on semantic analysis of the input sequence. 6. The method according to claim 5, wherein the method further comprises the following: - Carrying out the following steps respectively for at least one sudden increase keyword contained in the user demand sudden increase keyword library: - Obtain click information of one or more users on the sorted resource items obtained based on the sudden keyword or an input sequence containing the sudden keyword; - According to the click information, it is judged whether the timeliness of the sudden increase keyword has expired; - When it is judged that the timeliness of the sudden increase keyword has expired, delete the sudden increase keyword in the user demand sudden increase keyword database. 7. The method according to claim 4 or 6, wherein said b comprises the steps of: - judging whether the input sequence is time-sensitive according to whether the time-sensitive feature information satisfies a predetermined condition; - adding the input sequence and/or keywords in the input sequence to the user demand sudden increase keyword library when it is judged that the input sequence has timeliness. 8. The method according to claim 1, wherein the method further comprises the steps of: -according to the timeliness information, determine the change intensity of the user demand of the input sequence; - Determine the weighted value according to the change intensity of the user's demand; Wherein, the weighting processing steps include: - performing said weighting process on said at least one resource item according to said weighted value. 9. The method according to claim 8, wherein said weighting processing step further comprises: - performing the weighting process on the at least one resource item according to the weighted value and in combination with the establishment or update time of each resource item in the at least one resource item. 10. The method of claim 1, wherein the freshness-related information includes at least one of the following: - the creation or update time of the resource item and the predetermined user demand surge time of said input sequence; - resource items contained in the historical retrieval records of the input sequence. 11. A sorting device for sorting resource items, wherein the sorting device comprises: A timeliness information acquisition device, configured to acquire timeliness information related to the input sequence according to the input sequence from the user; The first judging means is used for judging whether the input sequence has timeliness according to whether the timeliness information satisfies a predetermined condition; A timeliness sorting device, configured to perform timeliness sorting on the resource items retrieved based on the input sequence when it is judged that the input sequence is timeliness, so as to provide the sorted resource items to the user; Wherein, the aging sorting device includes: The sub-sorting device is used to select at least one resource item from among them according to the freshness-related information of each resource item, and perform weighting processing on the selected at least one resource item, so as to sort according to the processing result after the weighting processing, Wherein, the freshness related information includes related information that can be used to determine whether the establishment or update time of each resource item belongs to the aging period of the input sequence. 12. The sorting device according to claim 11, wherein the aging information obtaining device comprises: The first sub-acquisition means is configured to acquire, according to the input sequence, time-sensitive characteristic information related to the input sequence as the time-sensitive information. 13. The sorting device according to claim 11, wherein the aging information obtaining device comprises: Inquiry means, for inquiring keywords related to the input sequence in the pre-established user demand sudden increase keyword library according to the input sequence; The defining means is used to use the query result as the time-sensitive information when the keyword related to the input sequence is found. 14. The sorting device according to claim 13, wherein the aging information obtaining device further comprises: The second sub-acquisition means is used to acquire, according to the input sequence, time-sensitive feature information related to the input sequence as the time-sensitive information when no keyword related to the input sequence can be found. 15. The sorting device according to claim 14, wherein the aging characteristic information includes at least one of the following: - information on changes in the quantity of resource items over time obtained by searching based on the input sequence; - information on the number of searches of the input sequence and/or keywords included in the input sequence over time; - burst semantic feature information of the input sequence obtained based on semantic analysis of the input sequence. 16. The sorting device according to claim 15, wherein the sorting device further comprises: An updating device, configured to perform operations on at least one sudden increase keyword included in the user demand sudden increase keyword library, wherein the update device includes: Click information acquiring means, configured to acquire click information of one or more users on sorted resource items obtained based on the sudden increase keyword or an input sequence containing the sudden increase keyword; The second judging means is used to judge whether the timeliness of the sudden increase keyword has expired according to the click information; The deletion device is used for deleting the sudden increase keyword from the user demand sudden increase keyword database when it is judged that the timeliness of the sudden increase keyword has expired. 17. The sorting device according to claim 14 or 16, wherein said first judging means comprises: A sub-determining device, configured to determine whether the input sequence is time-sensitive according to whether the time-sensitive feature information satisfies a predetermined condition; Adding means for adding the input sequence and/or keywords in the input sequence into the user demand sudden increase keyword library when it is judged that the input sequence has timeliness. 18. The sorting device according to claim 11, wherein the sorting device further comprises: Intensity determining means, configured to determine the intensity of user demand change in the input sequence according to the time-sensitive information; a weighted value determination device, configured to determine a weighted value according to the change intensity of the user's demand; Wherein, the sub-sorting device includes: Weighting processing means, configured to perform the weighting processing on the at least one resource item according to the weighting value. 19. The sorting device according to claim 18, wherein the weighting processing device further comprises: The sub-weighting processing means is configured to perform the weighting processing on the at least one resource item according to the weighting value and in combination with the establishment or update time of each resource item in the at least one resource item. 20. The sorting device according to claim 11, wherein the freshness-related information comprises at least one of the following: - the creation or update time of the resource item and the predetermined user demand surge time of said input sequence; - resource items contained in the historical retrieval records of the input sequence. 21. A computer device comprising a sorting device as claimed in any one of claims 11 to 20.