JP2011221591A - Map retrieval system - Google Patents

Abstract

An object of the present invention is to provide a map search and display device that can identify a candidate point where a reporter is present in a short time and with few operation steps without depending on the experience and intuition of the receptionist.
The speech recognition information converted into character data by the speech recognition processing device 1 and the target data in the database 8 of the map search device 6 are sequentially referred to, so that the words uttered by the reporter 2 or the receptionist 3 are expressed. In addition to automatically extracting the target data included, the first target data extracted from the extracted target data is used as the reference data, and the target data and reference data extracted from the second onward are within the specified values. We narrowed down the search for existing data and displayed the target reference data as candidate points.
[Selection] Figure 1

Description

  In the emergency call reception center such as No. 110, the present invention specifies a candidate point where the reporter is present from data obtained by characterizing the words uttered by the reporter and / or the receiver, and creates a map of the candidate point. The present invention relates to a map search system to be displayed.

In the conventional map search and display device, from the name of the nearby target feature heard from the whistleblower, the classification and name of the target based on the judgment of the receptionist, and a search menu on the screen prepared in advance By selecting an item, a list of candidate targets is displayed, and by searching for and selecting a target feature from the list, a map of the corresponding point is displayed on the terminal screen.
Further, as a conventional technique using voice recognition, there is a technique that displays a corresponding point map of a target feature that is a target by extracting only one target such as a name that completely matches. (For example, see Patent Document 1)

JP 2009-239466 A (FIGS. 8 and 9)

  In conventional map search and display devices, the task of extracting information that can be a target from information heard from the reporter is largely dependent on the experience and intuition of the receiver, and depending on the receiver, the location of the reporter It may take a long time to identify. In addition, the necessity of a multi-step terminal operation has also been a problem that it takes time to specify the location of the reporter.

  This invention has been made to solve the above-mentioned problems, and by adopting a voice recognition technology, the name, type, address, etc., can be obtained from data obtained by characterizing the words produced by the reporter or the receptionist. It is an object of the present invention to obtain a map search and display device that can automatically extract search information and identify candidate points where a reporter exists in a short time and with few operation steps.

  A map search system according to the present invention includes a speech recognition processing device that recognizes input words and converts them into character data, and a map search device that extracts target data from a speech recognition result and searches and displays a map of a corresponding point. The map search device includes a database storing target data such as names, types, and addresses, and sequentially refers to voice recognition information converted into character data by the voice recognition processing device and data in the database. In this way, the target data included in the input word is automatically extracted, and among the extracted target data, the first extracted target data is used as the reference data, and the target extracted after the second A search that narrows the object data and the reference data within a predetermined distance, and displays the target reference data as candidate points. A.

  According to the map search system of the present invention, search information such as name, type, address, etc. is automatically extracted from data obtained by characterizing the words of the reporter or the receptionist, and the reporter exists in a short time and with few operation steps. Candidate points can be identified.

It is a block diagram which shows the structure of Embodiment 1 of this invention. FIG. 3 is a flowchart of target data search processing according to the first embodiment. FIG. 10 is a diagram illustrating an example of an operation screen according to the third embodiment.

Embodiment 1 FIG.
A first embodiment of the present invention will be described below with reference to FIG.
In FIG. 1, a speech recognition processing device 1 includes a speech input unit 4 that inputs speech uttered by a reporter 2 and a receiver 3, a speech recognition processing unit 5 that converts speech into character data, and speech recognition that is converted into character data. It consists of a character data output unit 7 for handing over information to the map search device 6.
The map search device 6 includes a target database 8 prepared in advance, a reference processing unit 9 for collating each data of the speech recognition information and the target database 8, and a target obtained as a result of the collation processing by the reference processing unit 9. A target data search processing unit 10 for narrowing down and searching for target data serving as candidate points based on the object data, and a display monitor 11 for displaying a map of candidate points on the screen.

Next, the operation will be described.
In a phone call made from the reporter 2 to the acceptor 3 by the 110th report or the like, all utterance contents exchanged for confirming the situation are sequentially converted into character data in the speech recognition processing device 1 and become speech recognition information. Is output.
The voice recognition information is collated with each data stored in the target database 8 of the map search device 6, and all target data matching the voice recognition information is extracted. The target database 8 has a tabular data structure consisting of name, type, address, and coordinates, and the speech recognition information output each time an utterance is collated with each item of the target database 8, As the matched target data, first, second, third,... Are sequentially acquired.

Hereinafter, the operation of the target data search processing unit 10 will be described.
Among the target data that matches the voice recognition information, when the first voice recognition information is not a unique name but matches a general name or type, a plurality of target target data are extracted. This target object data group is set as reference data.
The example of the first embodiment shows a case where the reporter 2 receives a report “There is a school nearby. There is a convenience store and a post office-like building”. In this case, “school” is the reference data.

Details of the target data search will be described below with reference to FIG.
First, as described above, target data corresponding to the type “school” serving as reference data is extracted in step 1. Next, in step 2, target data corresponding to the next target data “convenience store” is extracted. Next, out of all the target data “school” extracted in step 1 in step 3, the target data “school” in which the target data “convenience store” extracted in step 2 is within a predetermined distance range is narrowed down. Extract with In step 4, target data corresponding to the next target data “post office” is extracted. Of all the target data “schools” extracted and refined in step 3 in step 5, the target “school” in which the target data “post office” extracted in step 4 is within a predetermined distance range is narrowed down. Extract. The above step 4 and step 5 are repeatedly performed (step 6), and the target data “school” of the reference data is narrowed down.

  As described above, each piece of data of the second target object data group is compared with the plurality of pieces of reference data extracted earlier, and the reference data whose distance from the reference data is within a predetermined value is narrowed down and searched. Thereafter, the search for the reference data is similarly performed for the third voice recognition information and the fourth voice recognition information. In the example of the first embodiment, the second target data group is “convenience store”, and the third target data group is “post office”. In this way, candidate reference data is narrowed down more and more.

When narrowing down to a plurality of cases, a candidate list is displayed on the screen, and by designating any one of them, a map of the corresponding point can be displayed on the display monitor 11 (step 7).
In addition, when the number is finally narrowed down to one, a map of the corresponding point is displayed on the display monitor 11 based on the coordinates of target target data.
By the above operation, the receptionist 3 can answer the phone call from the reporter 2 and listen to necessary information to automatically find a candidate point where the reporter 2 is located, thereby reducing the time required for specifying the location. In addition, the operability can be improved and the location identification accuracy can be improved.

Embodiment 2. FIG.
In the first embodiment, the operation when a plurality of target data groups exist within a predetermined distance range with respect to the reference data and the candidate points are found has been described. For each item, if the voice recognition information is information about an address, it corresponds as a step before performing the distance comparison processing between the reference data to be extracted and the plurality of target data groups described in the first embodiment. Search is limited to target data having address information. In this way, the number of data for comparing the distances between the target data can be narrowed down in advance, and the speed of the target data extraction process can be increased.

Embodiment 3 FIG.
In the first embodiment, all the data extracted as the speech recognition information is the target of the target data extraction process, and the operation in the case of finding the candidate point is described. However, as in the screen example shown in FIG. By making it possible to freely select whether or not to use individual target data as a narrowing-down condition in the speech recognition information, the degree of freedom for selecting candidate points can be expanded. In the example of FIG. 3, the reference data “school” and the target data “convenience store” are selected, but the target data “post office” is excluded and searched. By providing the degree of freedom in this way, it is possible to prevent omission of retrieval of target target data by relaxing the combination conditions when the target data is narrowed down too much.

  DESCRIPTION OF SYMBOLS 1 Voice recognition processing device, 2 Reporter, 3 Receiving person, 4 Voice input part, 5 Voice recognition processing part, 6 Map search device, 7 Character data output part, 8 Target object database, 9 Reference processing part, 10 Target object data Search processing unit, 11 display monitor.

Claims (3)

  A speech recognition processing device that recognizes an input word and converts it into character data, and a map search device that extracts target data from the speech recognition result and searches and displays a map of the corresponding point. A database that stores target data such as names, types, addresses, etc., and the speech recognition information that has been converted into character data by the speech recognition processing device, and the data entered in the database by sequentially referring to the data in the database The target data included in the is automatically extracted, and the first target data extracted from the extracted target data is used as the reference data, and the second and subsequent target data and reference data are set as default values. A map search system that narrows down the search within the distance and displays the target reference data as candidate points.   When the target data extracted by the voice recognition processing device is information related to an address, the search is limited to target data existing in the corresponding address area as a candidate point narrowing method. The map search system according to claim 1.   In the map search device, whether or not to use each target data as a narrowing condition can be freely selected for a plurality of target data obtained by the sequential reference processing of the speech recognition information and target data. The map search system according to claim 1 or 2, wherein a candidate point can be searched based on a combination of the target data.

Images