HK1192035A

HK1192035A - Method for providing search result, and apparatus thereof

Info

Publication number: HK1192035A
Application number: HK14105111.7A
Authority: HK
Inventors: 李嘉森; 姚建强; 帅朝谦
Original assignee: 阿里巴巴集团控股有限公司
Filing date: 2014-05-30
Publication date: 2014-08-08

Description

Method and device for providing search results

Technical Field

The present application relates to the field of communications technologies, and in particular, to a method and an apparatus for providing a search result.

Background

Currently, most shopping websites can provide a search function for a user, and the user can search commodity information of a commodity desired by the user through the search function. Specifically, the user sends a search word to a server of the shopping website, and the server searches for the commodity information related to the search word based on the received search word and provides the searched commodity information to the user.

In practical applications, since the user usually focuses on the commodity information ranked at the top, after the server searches for the commodity information, the server needs to rank the searched commodity information and provide the ranked commodity information to the user. However, since a general user pays attention to prices included in the commodity information, in the prior art, when the server sorts the searched commodity information, the searched commodity information may be sorted in order of the included prices from high to low or from low to high.

However, the selling prices of the same product made by different sellers are different for the same product. For example, the search term transmitted from the user as the buyer to the server is a certain brand of running shoes, and among the product information searched by the server, the price included in 10 product information is 20 yuan, the price included in 20 product information is 100 yuan, and the price included in 5 product information is 300 yuan. While the price of the running shoes of the brand should be 120 yuan in practice, obviously, the user as the buyer can intuitively think that the running shoes are sold in 100 yuan more reasonably, so as to pay attention to the 20 commodity information with the price of 100 yuan.

However, in the prior art, the server can only sort the searched commodity information according to the order from high price to low price or from low price to high price, and if the server sorts the commodity information searched in the above example according to the order from low price to high price, the sorting result is: 10 items of contained commodity information with a price of 20 yuan, 20 items of contained commodity information with a price of 100 yuan, and 5 items of contained commodity information with a price of 300 yuan.

Obviously, the server does not arrange the 20 items of commodity information with the price of 100 yuan at the front position, but arranges the 10 items of commodity information with the price of 20 yuan at the front position, so that after the server provides the sorted commodity information for the user, the commodity corresponding to the sorted commodity information at the front position is not a commodity which the user considers reasonable in price, and the user can continuously set a screening condition for re-searching, thereby increasing the pressure of the server.

Disclosure of Invention

The embodiment of the application provides a method and a device for providing a search result, which are used for solving the problem that in the prior art, after a server provides sorted commodity information for a user, the user still needs to continuously set a screening condition for re-searching, so that the server is stressed greatly.

The method for providing the search result provided by the embodiment of the application comprises the following steps:

the server searches related commodity information based on the received search words; and are

Searching a price optimal value corresponding to the predetermined search word; and

and respectively determining the absolute value of the difference value between the price contained in each searched commodity information and the found optimal value of the price, and sequencing and providing the searched commodity information according to the sequence of the respectively determined absolute values from small to large.

The device for providing the search result provided by the embodiment of the application comprises:

the search module is used for searching related commodity information based on the received search words;

the optimal value determining module is used for searching a predetermined optimal value of the price corresponding to the search term;

and the providing module is used for respectively determining the absolute value of the difference value between the price contained in the searched commodity information and the found optimal value of the price, sequencing the searched commodity information according to the sequence of the respectively determined absolute values from small to large and providing the sequenced commodity information.

The embodiment of the application provides a method and a device for providing search results, and the method comprises the steps that when a server searches based on received search terms, a predetermined optimal price value corresponding to the search terms is searched, the absolute value of the difference value between the price contained in each piece of searched commodity information and the optimal price value is respectively determined, and the searched commodity information is sequenced and provided according to the determined absolute value from small to large. Through the method, the server provides the commodity information with the commodity information in the top order, namely the commodity information with the smaller difference between the contained price and the optimal price value, and compared with the method in the prior art that the search results are provided singly according to the sequence of the contained prices from high to low or from low to high, the method can effectively reduce the times of setting the screening condition and searching again by the user, thereby reducing the pressure of the server.

Drawings

FIG. 1 is a process for providing search results provided by an embodiment of the present application;

fig. 2 is a process of fitting a probability density function based on prices included in the commodity information searched by the search term according to an embodiment of the present application;

fig. 3 is a schematic structural diagram of an apparatus for providing a search result according to an embodiment of the present application.

Detailed Description

Because the server in the prior art can only sort and provide the searched commodity information according to the order of the included prices from high to low or from low to high, the prices included in the commodity information with the top sorting order provided by the server in the prior art are often far from the reasonable prices considered by most users, which causes the users to continuously set the screening conditions for re-searching, and increases the pressure of the server.

The server determines the optimal price value corresponding to the search word in advance, the optimal price value corresponding to the search word is a reasonable price considered by most users in prices contained in the commodity information which can be searched based on the search word, and when searching is carried out based on the search word, the searched commodity information is ranked and provided in a mode that the ranking is more forward as the absolute value of the difference between the price contained in the searched commodity information and the optimal price value corresponding to the search word determined in advance is smaller, so that the commodity information which is provided to the users and ranked earlier is commodity information which contains prices closer to the optimal price value corresponding to the search word, the frequency of searching again by the users through setting screening conditions is reduced, and the pressure of the server is reduced.

The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

Fig. 1 is a process for providing a search result according to an embodiment of the present application, which specifically includes the following steps:

s101: the server searches for related goods information based on the received search word.

Similar to the prior art, when a user wants to search commodity information of a certain commodity, a corresponding search word is sent to a server, and after receiving the search word sent by the user, the server searches the commodity information related to the search word based on the received search word.

S102: and searching a preset optimal value of the price corresponding to the search word.

In the embodiment of the application, the server determines the corresponding optimal price value for each search term recorded in the search log in advance, so that the server searches for the predetermined optimal price value corresponding to the search term after searching based on the received search term.

The method for predetermining the optimal price value corresponding to the search term may be: and according to prices contained in all the commodity information which can be searched based on the search word, taking the average price of the prices contained in all the commodity information as the optimal price value corresponding to the search word, or taking the median of the prices contained in all the commodity information as the optimal price value corresponding to the search word. Of course, the optimal price value corresponding to the search term may also be set manually.

S103: and respectively determining the absolute value of the difference value between the price contained in each searched commodity information and the found optimal value of the price, and sequencing and providing the searched commodity information according to the sequence of the respectively determined absolute values from small to large.

That is, when the searched commodity information is sorted, the sorted commodity information is sorted in a manner that the smaller the absolute value of the difference between the included price and the optimal price value is, the more the sorted commodity information is, because the optimal price value corresponding to the predetermined search word is closer to the reasonable price of the commodity searched by the search word, which is considered by most users, the user can effectively reduce the number of times that the user sets the screening condition to re-search after sorting the searched commodity information according to the sorting manner in step S103 and providing the sorted commodity information to the user, thereby reducing the pressure of the server.

In the embodiment of the present application, the server needs to determine a corresponding optimal price value for each used search term recorded in the search log in advance, and the method for determining the optimal price value corresponding to a certain search term specifically includes that the server searches for commodity information that satisfies a specified condition in the searched commodity information every time the server searches based on the search term in the past recorded in the search log, and determines the optimal price value corresponding to the search term according to the price included in the searched commodity information. The commodity information satisfying the specified condition includes commodity information clicked by the user. Of course, the commodity information satisfying the specified condition may further include commodity information collected by the user, commodity information in which a corresponding commodity is ordered, shared commodity information, and commodity information in which other users have performed a specified operation. In addition, the server may search for commodity information satisfying a predetermined condition among the commodity information searched for each time a search is performed based on the search term in a predetermined time period in the past when searching for commodity information satisfying the predetermined condition every time a search is performed based on the search term in the past.

In practical applications, after the server provides the commodity information searched based on the search term to the user, the user usually pays attention to the commodity information considered to be included with a reasonable price, and for the commodity information considered to be included with a reasonable price, the user usually performs a click operation to view the commodity information in detail, while for the commodity information considered to be included with an unreasonable price, the user usually does not perform a click operation to ignore the commodity information. Therefore, the server determines the optimal value of the price corresponding to the search term based on the price included in the clicked item information in the item information searched for based on the search term in the past.

For example, assuming that the optimal price value corresponding to the search word "1 g display card" is to be determined, the server searches the search log for the past search behavior performed by using "1 g display card" as the search word. Assuming that 3 searching behaviors are found, further finding commodity information meeting specified conditions in the commodity information searched each time when the 3 searching behaviors are carried out, namely finding commodity information clicked by the user in the searched commodity information after searching by taking the '1 g display card' as a search word each time. Assuming that the clicked commodity information in the commodity information searched for at the 1 st time is information of a commodity A, the clicked commodity information in the commodity information searched for at the 2 nd time is information of a commodity B, and the clicked commodity information in the commodity information searched for at the 3 rd time is information of a commodity C, the server determines the optimal price value corresponding to the search word "1 g display card" according to the price contained in the searched information of the commodity A, the price contained in the information of the commodity B and the price contained in the information of the commodity C. The average value of the prices included in the searched commodity information may be determined as the optimal price value corresponding to the search term, that is, the average values of the prices included in the information of the commodity a, the information of the commodity B, and the information of the commodity C are used as the optimal price values corresponding to the search term.

Of course, when the optimal value of the price corresponding to the search term is determined according to the price included in the searched commodity information, the price included in the searched commodity information can be converted into a unified measurement price according to the actual situation. For example, the price contained in the searched commodity information is converted into the price of each commodity according to the actual situation, or the price of each kilogram of commodities, or the price of each liter of commodities, or the price of each square meter of commodities, and the like.

By the method, the server can more accurately determine the optimal price value corresponding to the search term, so that when searching is performed subsequently based on the search term, after the searched commodity information is sequenced and provided through the step S103 shown in FIG. 1, the frequency of re-searching under the condition that the user sets the screening condition can be further reduced, and the pressure of the server can be further relieved.

Considering that there are many similar search terms in practical application, such as "1 g display card" and "display card 1 g", when two search terms are used for searching, the searched commodity information is basically the same, so the server can sort the search terms recorded in the search log and integrate the similar search terms into one normalized search term, for example, integrate the search term "1 g display card" and the search term "display card 1 g" into the normalized search term "1 g display card". When commodity information meeting specified conditions in the commodity information searched by the search word in the past is searched, searching is carried out in the search log based on the normalized search word of the search word.

For example, after the search word "1 g display card" and the search word "display card 1 g" recorded in the search log are integrated into the normalized search word "1 g display card", if the optimal price value corresponding to the search word "display card 1 g" is to be determined, the normalized search word of the search word "display card 1 g" is determined to be "1 g display card", the commodity information clicked by the user, which is searched each time the search is performed based on the "1 g display card" in the past and recorded in the search log, is searched, and the optimal price value corresponding to the search word "display card 1 g" is determined according to the price included in the searched commodity information.

Preferably, in practical applications, for the commodity information searched by using the same search word, the probability density distribution of the prices included in each searched commodity information approximately conforms to the mixed double-gaussian probability distribution, so that when the server determines the optimal value of the price corresponding to the search word according to the prices included in each searched commodity information meeting the specified conditions, the server may adopt a mixed double-gaussian model to fit a probability density function based on the prices included in the commodity information searched by the search word according to the prices included in the searched commodity information, and determine the price corresponding to the maximum probability density as the optimal value of the price corresponding to the search word according to the fitted probability density function.

Fig. 2 is a specific diagram illustrating a process of fitting a probability density function of a price included in commodity information searched based on a search term by using a hybrid double-gaussian model, where fig. 2 is a process of fitting a probability density function of a price included in commodity information searched based on a search term provided in an embodiment of the present application, and the process specifically includes the following steps:

s201: and determining the total average value of the prices contained in the searched commodity information according to the prices contained in the searched commodity information.

For example, to determine the optimal price value corresponding to the search term "1 g display card", when the server searches for a search line that was performed using the "1 g display card" as the search term in the past in the search log, the server searches for product information that satisfies the specified condition among the searched product information. Suppose that n commodity information meeting specified conditions are found, wherein the price contained in the ith commodity information is recorded as x_iThen determining the n searched commodity informationTotal average value of included prices

S202: and determining the total standard deviation of the prices contained in the searched commodity information according to the prices contained in the searched commodity information and the determined total average value.

Continuing with the above example, the total average of the prices contained in the n item information is determinedThen, the formula is adoptedThe total standard deviation σ of the prices contained in the n commodity information is determined.

Preferably, for the prices included in the searched commodity information, if the price included in a certain commodity information is greatly different from the determined total average value, the price included in the commodity information is abnormal data, and the abnormal data affects the accuracy of subsequent fitting. Specifically, continuing with the above example, the price x included in the ith item information of the n item information found is pointed to_iIf, ifIf the value of (b) is greater than 3, the price x included in the ith commodity information is determined_iEliminating the price x contained in the ith commodity information as abnormal data_i. Or, ifIf the value of (b) is not within the set range, the price x included in the ith item information is determined_iEliminating the price x contained in the ith commodity information as abnormal data_iThe set range may be, for example, 0.2 to 1.8.

S203: and taking the product of the determined total average value and the first setting parameter as a first average value, and taking the product of the determined total average value and the second setting parameter as a second average value, wherein the first setting parameter is more than 1, and the second setting parameter is less than 1.

Continuing with the example above, the first average value determinedDetermined second mean valueIn the embodiment of the present application, it is only required to ensure that one of a and b is greater than 1 and the other is less than 1, for example, a is 0.3 and b is 1.3.

S204: and simultaneously taking the determined total standard deviation as the first standard deviation and the second standard deviation.

Continuing with the above example, the first standard deviation σ is determined₁σ, second standard deviation σ determined₂＝σ。

S205: aiming at the price contained in each searched commodity information, a formula is adoptedAnd determining a middle value corresponding to each price.

Wherein, p is the weight value of the initial setting. The weight value of the initial setting may be set to 0.5.

Continuing with the above example, the price x included in the ith commodity information of the found n commodity information is pointed out_iDetermining the price x contained in the ith commodity information by adopting the formula_iCorresponding intermediate value gamma (i).

S206: according to the determined intermediate value corresponding to each price, adopting a formulaRe-determining the first average value by using the formulaThe second average is re-determined.

S207: according to the re-determined first average value and the second average value, adopting a formulaRe-determining the first standard deviation by using the formulaThe second standard deviation is redetermined.

Wherein, the formulaAndmu in₁And mu₂The first average value and the second average value re-determined in step S206.

S208: using a formulaAnd re-determining the weight value.

Wherein n is the number of the searched commodity information.

S209: and judging whether the redetermined first average value, second average value, first standard deviation and second standard deviation meet set conditions, if so, executing step S211, otherwise, executing step S210.

In the embodiment of the present application, the method for determining whether the redetermined first average value, second average value, first standard deviation, and second standard deviation satisfy the set condition specifically includes: judging that the difference value of the redetermined first average value and the last determined first average value is in a set range, judging that the difference value of the redetermined second average value and the last determined second average value is in a set range, judging that the difference value of the redetermined first standard deviation and the last determined first standard deviation is in a set range, and judging that the difference value of the redetermined second standard deviation and the last determined second standard deviation is in a set range; and when at least one judgment result in the four judgments is yes, judging whether the redetermined first average value, second average value, first standard deviation and second standard deviation meet set conditions.

Of course, when at least two or all of the four determinations are yes, it may be determined whether the redetermined first average value, second average value, first standard deviation and second standard deviation satisfy the set condition.

S210: and re-determining the intermediate value corresponding to each price according to the re-determined first average value, second average value, first standard deviation, second standard deviation and weight, and returning to the step S206.

That is, based on the re-determined first average value, second average value, first standard deviation, second standard deviation, weight, formula is adoptedAnd re-determining the intermediate value corresponding to the price contained in each piece of searched commodity information, and continuously determining the first average value, the second average value, the first standard deviation, the second standard deviation and the weight according to each re-determined intermediate value until the determined first average value, second average value, first standard deviation and second standard deviation meet the set conditions.

S211: based on the re-determined first average valueDetermining a function of the mean value, the first standard deviation, the second standard and the weightAnd as a fitted probability density function of prices contained in the commodity information searched based on the search word.

Continuing to use the above example, when the first average value, the second average value, the first standard deviation and the second standard deviation which meet the set conditions are obtained, the function is obtained according to the first average value, the second average value, the first standard deviation and the second standard deviation which meet the set conditions and the finally determined weightNamely, the fitted probability density function of the prices contained in the commodity information searched based on the search word "1 g display card".

In the subsequent process, the price corresponding to the maximum probability density can be determined as the optimal price value corresponding to the search term "1 g display card", that is, the obtained function takes the maximum function value f (x)_maxThe value of the corresponding argument x is the optimal value of the price corresponding to the search term "1 g display card". The probability density function is obtained based on the price contained in the searched commodity information meeting the specified condition, and the commodity information meeting the specified condition is the commodity information clicked by the user, so that the probability that the commodity information searched by the user is clicked by the user is reflected by the probability density function, the probability that the commodity information containing a certain price is clicked is higher, the more users think that the price is a reasonable price is, the price corresponding to the highest probability density is taken as the optimal price value corresponding to the search word, the optimal price value corresponding to the search word can be determined more accurately, the frequency of re-searching by setting the screening condition by the user is further reduced, and the pressure of the server is reduced.

Wherein the probability density function of the price contained in the commodity information searched based on the search term is obtainedSince then the maximum value of the function must be at mu₁And mu₂Thus determining the function as the maximum function value f (x)_maxSelf-corresponding of timeThe method of the value of the variable x may be: mu to mu₁To mu₂Is divided into several sub-ranges, e.g. mu may be₁To mu₂Is divided into n sub-ranges, each being mu₁To mu₁+d、mu₁+ d to mu₁+2d、mu₁+2d to mu₁+3d……、mu₁D to mu (n-1)₂Wherein, in the step (A),randomly selecting a value of x in each sub-range of the division according to the obtained probability density functionDetermining the function value f (x) corresponding to each selected value of x, determining the value of x corresponding to the maximum function value in each determined function value f (x) as the maximum function value f (x) of the function_maxThe value of the corresponding argument x, that is, the price optimization function corresponding to the search term.

Preferably, in the statistics, if it is determined that the probability density distribution of some samples conforms to the mixed double-gaussian probability distribution, the number of different samples is required to be greater than a certain threshold, so in the embodiment of the present application, before the process shown in fig. 2 is used to fit the probability density function of the prices included in the commodity information searched based on the search term, it is further determined that the number of the searched commodity information satisfying the specified condition and including different prices is greater than the set threshold.

In addition, considering that prices included in the product information searched based on some search terms are often high in practical application, the amount of calculation required when the process shown in fig. 2 is adopted to fit the probability density function of the prices included in the product information searched based on the search terms is large, and therefore, in step S201, the logarithm operation may be performed on the prices included in each piece of searched product information, and the obtained result may be used as the prices included in each piece of product information again.

For example, for the price x contained in the ith commodity information in the found n commodity information satisfying the specified condition_iLog may be obtained₂x_iThe value of (a) is re-used as the price contained in the ith commodity information, and then the fitting process shown in fig. 2 is performed according to the re-determined price contained in each commodity information. Assuming that the optimal price value corresponding to the search term is finally determined to be X according to the process shown in fig. 2, 2 is further added^XAnd the price is used as the determined optimal value corresponding to the search term again.

After the optimal price value corresponding to the search word is obtained, the obtained optimal price value of the search word can be manually adjusted, and the optimal price value of the search word is stored, so that the optimal price value of the search word can be searched when searching is carried out on the basis of the search word in the following, and searched commodities are ranked and provided according to prices contained in the searched commodity information and the optimal price value corresponding to the search word. Of course, when the server ranks the searched commodity information in the embodiment of the present application, in addition to ranking according to the price included in each searched commodity information and the optimal value of the price corresponding to the search term, the server may rank the searched commodity information by integrating the correlation between the searched commodity information and the search term, the quality of the commodity corresponding to the searched commodity information, the quality of the seller corresponding to the searched commodity information, and other factors.

The server may also rank and provide the searched commodity information according to the price included in each searched commodity information and the price optimal value corresponding to the search word only when a request for ranking according to the price optimal value sent by the user is received, and rank and provide the searched commodity information according to a default ranking rule when the request for ranking according to the price optimal value is not received.

Fig. 3 is a schematic structural diagram of an apparatus for providing a search result according to an embodiment of the present application, which specifically includes:

a search module 301, configured to search for related commodity information based on the received search term;

an optimal value determining module 302, configured to search a predetermined optimal value of a price corresponding to the search term;

and the providing module 303 is configured to determine absolute values of differences between the prices included in each piece of searched commodity information and the found optimal price value, sort and provide the searched commodity information according to a descending order of the determined absolute values.

The optimal value determining module 302 is specifically configured to search for commodity information that satisfies a specified condition in the commodity information that is searched for each time in the past when searching is performed based on the search term, where the commodity information that satisfies the specified condition includes commodity information clicked by the user; and determining the optimal price value corresponding to the search word according to the price contained in the searched commodity information.

The optimal value determining module 302 is specifically configured to, according to the prices included in the searched commodity information, fit a probability density function based on the prices included in the commodity information searched by the search term by using a hybrid double-gaussian model, and according to the fitted probability density function, determine a price corresponding to the maximum probability density as the optimal value of the price corresponding to the search term.

The optimal value determining module 302 specifically includes:

an initialization unit 3021, configured to determine a total average value of prices included in the searched commodity information according to prices included in the searched commodity information, determine a total standard deviation of prices included in the searched commodity information according to prices included in the searched commodity information and the determined total average value, take a product of the determined total average value and a first setting parameter as a first average value, and take a product of the determined total average value and a second setting parameter as a second average value, where the first setting parameter is greater than 1, and the second setting parameter is less than 1; taking the determined total standard deviation as a first standard deviation and a second standard deviation at the same time;

training unit 3022 for findingThe price contained in each commodity information adopts a formulaDetermining a middle value corresponding to each price, wherein p is an initially set weight value, mu₁Is a first mean value, mu₂Is the second mean value, σ₁Is the first standard deviation, σ₂Is the second standard deviation, x_iFor the price contained in the searched ith commodity information, gamma (i) is a middle value corresponding to the price contained in the searched ith commodity information; according to the determined intermediate value corresponding to each price, adopting a formulaRe-determining the first average value by using the formulaRe-determining the second average value; according to the re-determined first average value and the second average value, adopting a formulaRe-determining the first standard deviation by using the formulaRe-determining the second standard deviation; using a formulaRe-determining the weight value, wherein n is the number of the searched commodity information; judging whether the redetermined first average value, second average value, first standard deviation and second standard deviation meet set conditions or not; when the judgment result is negative, re-determining the intermediate value corresponding to each price according to the re-determined first average value, second average value, first standard deviation, second standard deviation and weight, and continuously determining the first average value, second average value, first standard deviation, second standard deviation and weight according to the re-determined intermediate values until the determined first average value, second average value, first standard deviation and second standard deviation meet the set condition;

a function determining unit 3023, configured to determine a function according to the redetermined first average value, second average value, first standard deviation, and weight when the training unit 3022 determines that the redetermined first average value, second average value, first standard deviation, and second standard deviation satisfy the setting conditionAs a fitBased on a probability density function of prices contained in the commodity information searched for by the search term.

The training unit 3022 is specifically configured to determine that the redetermined first average value, the second average value, the first standard deviation and the second standard deviation satisfy the setting condition when it is determined that a difference between the redetermined first average value and the last determined first average value is within a setting range, or when it is determined that a difference between the redetermined second average value and the last determined second average value is within a setting range, or when it is determined that a difference between the redetermined second standard deviation and the last determined second standard deviation is within a setting range.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims

1. A method of providing search results, comprising:

2. The method of claim 1, wherein determining the optimal value of the price corresponding to the search term comprises:

the server searches commodity information which meets specified conditions in the searched commodity information every time when searching is carried out on the basis of the search words in the past and recorded in a search log, wherein the commodity information meeting the specified conditions comprises commodity information clicked by a user;

and determining the optimal price value corresponding to the search word according to the price contained in the searched commodity information.

3. The method according to claim 2, wherein determining the optimal value of the price corresponding to the search term according to the price included in the searched commodity information specifically comprises:

according to the price contained in the searched commodity information, fitting a probability density function based on the price contained in the commodity information searched by the search word by adopting a mixed double-Gaussian model;

and determining the price corresponding to the maximum probability density as the optimal price value corresponding to the search word according to the fitted probability density function.

4. The method according to claim 3, wherein fitting a probability density function based on prices included in the commodity information searched for by the search term using a mixture double-gaussian model specifically comprises:

determining a total average value of prices contained in the searched commodity information according to prices contained in the searched commodity information;

determining the total standard deviation of the prices contained in the searched commodity information according to the prices contained in the searched commodity information and the determined total average value;

taking the product of the determined total average value and a first set parameter as a first average value, and taking the product of the determined total average value and a second set parameter as a second average value, wherein the first set parameter is greater than 1, and the second set parameter is less than 1;

taking the determined total standard deviation as a first standard deviation and a second standard deviation at the same time;

aiming at the price contained in each searched commodity information, a formula is adoptedDetermine eachThe intermediate value corresponding to each price, wherein p is the weight value of the initial setting, mu₁Is a first mean value, mu₂Is the second mean value, σ₁Is the first standard deviation, σ₂Is the second standard deviation, x_iFor the price contained in the searched ith commodity information, gamma (i) is a middle value corresponding to the price contained in the searched ith commodity information;

according to the determined intermediate value corresponding to each price, adopting a formulaRe-determining the first average value by using the formulaRe-determining the second average value;

according to the re-determined first average value and the second average value, adopting a formulaRe-determining the first standard deviation by using the formulaRe-determining the second standard deviation;

using a formulaRe-determining the weight value, wherein n is the number of the searched commodity information;

judging whether the redetermined first average value, second average value, first standard deviation and second standard deviation meet set conditions or not;

if yes, determining a function according to the re-determined first average value, second average value, first standard deviation, second standard and weightA probability density function of prices contained in the commodity information searched based on the search term as a fit;

otherwise, re-determining the intermediate value corresponding to each price according to the re-determined first average value, second average value, first standard deviation, second standard deviation and weight, and continuously determining the first average value, second average value, first standard deviation, second standard deviation and weight according to the re-determined intermediate values until the determined first average value, second average value, first standard deviation and second standard deviation meet the set conditions.

5. The method of claim 4, wherein determining that the re-determined first average, second average, first standard deviation, and second standard deviation satisfy the set condition comprises:

judging that the difference value between the redetermined first average value and the last confirmed first average value is in a set range; or

Judging that the difference value between the second average value determined again and the second average value determined last time is in a set range; or

Judging that the difference value of the redetermined first standard deviation and the last determined first standard deviation is in a set range; or

And judging that the difference value of the second standard deviation determined again and the second standard deviation determined last time is in the set range.

6. An apparatus for providing search results, comprising:

and the providing module is used for respectively determining the absolute value of the difference value between the price contained in each searched commodity information and the found optimal value of the price, sequencing the searched commodity information according to the sequence of the respectively determined absolute values from small to large and providing the sequenced commodity information.

7. The apparatus according to claim 6, wherein the optimal value determining module is specifically configured to search for commodity information that satisfies a specified condition among the commodity information searched each time a search is performed based on the search term in the past, which is recorded in the search log, wherein the commodity information that satisfies the specified condition includes commodity information clicked by the user; and determining the optimal price value corresponding to the search word according to the price contained in the searched commodity information.

8. The apparatus according to claim 7, wherein the optimal value determining module is specifically configured to fit a probability density function based on prices included in the commodity information searched for by the search term using a hybrid double gaussian model according to the prices included in the searched commodity information, and determine, according to the fitted probability density function, a price corresponding to the maximum probability density as the optimal value of the price corresponding to the search term.

9. The apparatus according to claim 8, wherein the optimal value determining module specifically includes:

an initialization unit, configured to determine a total average value of prices included in the searched commodity information according to prices included in the searched commodity information, determine a total standard deviation of the prices included in the searched commodity information according to the prices included in the searched commodity information and the determined total average value, take a product of the determined total average value and a first setting parameter as a first average value, and take a product of the determined total average value and a second setting parameter as a second average value, where the first setting parameter is greater than 1, and the second setting parameter is less than 1; taking the determined total standard deviation as a first standard deviation and a second standard deviation at the same time;

a training unit for adopting a formula according to the price contained in each searched commodity informationDetermining a middle value corresponding to each price, wherein p is an initially set weight value, mu₁Is a first mean value, mu₂Is the second mean value, σ₁Is the first standard deviation, σ₂Is the second standard deviation, x_iFor the price contained in the searched ith commodity information, gamma (i) is a middle value corresponding to the price contained in the searched ith commodity information; according to the determined intermediate value corresponding to each price, adopting a formulaRe-determining the first average value by using the formulaRe-determining the second average value; according to the re-determined first average value and the second average value, adopting a formulaRe-determining the first standard deviation by using the formulaRe-determining the second standard deviation; using a formulaRe-determining the weight value, wherein n is the number of the searched commodity information; judging whether the redetermined first average value, second average value, first standard deviation and second standard deviation meet set conditions or not; if the judgment result is negative, the second average value, the first standard deviation, the second standard deviation and the weight are determined againNewly determining a middle value corresponding to each price, and continuously determining a first average value, a second average value, a first standard deviation, a second standard deviation and a weight according to each newly determined middle value until the determined first average value, second average value, first standard deviation and second standard deviation meet the set conditions;

a function determining unit, configured to determine a function according to the redetermined first average value, second average value, first standard deviation, second standard deviation and weight when the training unit determines that the redetermined first average value, second average value, first standard deviation and second standard deviation satisfy the setting conditionAnd as a fitted probability density function of prices contained in the commodity information searched based on the search word.

10. The apparatus according to claim 9, wherein the training unit is specifically configured to determine that the first re-determined average, the second re-determined average, the first standard deviation, and the second standard deviation satisfy the setting condition when a difference between the first re-determined average and a first previously determined average is determined to be within a setting range, or when a difference between the second re-determined average and a second previously determined average is determined to be within a setting range, or when a difference between the first re-determined standard deviation and a second previously determined standard deviation is determined to be within a setting range.