JP2017228114A - Query analysis device, query analysis method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a query analysis device, a query analysis method and a program capable of analyzing needs of a user who performs search and/or change of the user's needs over time using queries.SOLUTION: A query analysis device comprises: a query information acquisition unit for acquiring query information in which a searched query and a time when the query was searched are associated for each user; an overlap level score calculation unit for, based on the query information, calculating an overlap level score indicating a degree of overlapping of a user who searched a first query and a user who searched a second query; a search time difference calculation unit for, based on the query information, calculating a search time difference which is a difference between a time when the first query was searched and a time when the second query was searched; and a display information generation unit for generating information for displaying the overlap level score and the search time difference in association with the first query and the second query.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a query analysis device, a query analysis method, and a program.

  Conventionally, a technique for creating a thesaurus dictionary based on a query input to a search site has been used. Specifically, a pair search word is generated by pairing a search word with a search time earlier and a search word with a later search time from a search word having a search time interval within a predetermined time. There is known a technique for generating a thesaurus dictionary by using (see Patent Document 1).

JP 2013-109701 A

  However, although the technique disclosed in Patent Document 1 can perform a search word conversion process using the generated thesaurus dictionary, it is possible to detect changes in user needs over time and user needs over time. Could not be analyzed using the query entered in.

  The present invention has been made in view of such circumstances, and a query analysis apparatus and query analysis that can analyze a user's needs to be searched and changes in the user's needs over time using a query. An object is to provide a method and a program.

  One aspect of the present invention is a query information acquisition unit that acquires query information in which a searched query and a time when the query is searched are associated for each user, and the query acquired by the query information acquisition unit. Based on the information, acquired by the query information acquisition unit, the redundancy score calculation unit that calculates the redundancy score indicating the degree of overlap between the user who searched the first query and the user who searched the second query A search time difference calculating unit that calculates a search time difference that is a difference between a time when the first query is searched and a time when the second query is searched based on the query information that has been searched; The redundancy score calculated by the score calculation unit and the search time difference calculated by the search time difference calculation unit in the first query and the second query. A display information generator for generating information for display with continuous, a query analyzing device comprising a.

  According to one embodiment of the present invention, it is possible to analyze a user's needs to be searched and changes in the user's needs over time using a query.

1 is a diagram illustrating a configuration of a query analysis system 10 according to an embodiment. It is a sequence diagram which shows the search process of the web page which concerns on embodiment. It is a figure which shows an example of the query information memorize | stored in the memory | storage part 120 which concerns on embodiment. It is a figure for demonstrating the calculation process of the duplication degree score which concerns on embodiment. It is a figure for demonstrating the calculation process of the search time difference which concerns on embodiment. It is a figure which shows an example of the query analysis window W before the analysis start which concerns on embodiment. It is a figure showing an example of query analysis window W after the end of analysis concerning an embodiment. It is a figure for demonstrating the clustering process which concerns on embodiment. It is a figure which shows an example of the heat map M which concerns on embodiment. It is a flowchart which shows the query analysis process which concerns on embodiment.

  Hereinafter, embodiments of a query analysis device, a query analysis method, and a program will be described with reference to the drawings. The query analysis device acquires a history of searches performed via a network or the like, extracts one or more second queries having a high correlation with the first query, and extracts the first query and the second query. It is a device that visualizes the relationship. The query analysis apparatus may be realized by installing a tool (program) in a computer, or may be an apparatus that provides an analysis result by a cloud service. The query analysis device can analyze the needs of users to be searched and the changes in user needs over time.

<1. Configuration of query analysis system>
FIG. 1 is a diagram illustrating a configuration of a query analysis system 10 according to the embodiment. The query analysis system 10 according to the embodiment includes a web server 100, a query analysis device 200, and a user terminal 300.

  Web server 100, query analysis device 200, and user terminal 300 are connected to network NW. The network NW includes, for example, a part or all of a wide area network (WAN), a local area network (LAN), the Internet, a provider device, a wireless base station, a dedicated line, and the like.

  The web server 100 includes a control unit 110 and a storage unit 120. For example, the control unit 110 may be realized by a processor of the web server 100 executing a program, LSI (Large Scale Integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or the like. It may be realized by hardware, or may be realized by cooperation of software and hardware.

  The storage unit 120 is realized by, for example, a random access memory (RAM), a read only memory (ROM), a hard disk drive (HDD), a flash memory, or a hybrid storage device in which a plurality of these are combined. Further, part or all of the storage unit 120 may be an external device accessible by the web server 100, such as NAS (Network Attached Storage) or an external storage server.

  The query analysis apparatus 200 is an apparatus used by an analyst who analyzes a query, and is a notebook computer or a desktop computer, but is not limited thereto. For example, the query analysis device 200 may be a mobile phone such as a smartphone, a tablet terminal, or a PDA (Personal Digital Assistant).

  The query analysis device 200 includes an input unit 210, a display unit 220, a query information acquisition unit 230, a redundancy score calculation unit 240, a search time difference calculation unit 250, a clustering unit 260, a heat map generation unit 270, A display information generation unit 280 and a storage unit 290 are provided. The input unit 210 is an input device such as a keyboard or a mouse. When the query analysis device 200 is a mobile phone such as a smartphone or a tablet terminal, the input unit 210 may be an input device such as a touch panel. The display unit 220 is a display device such as a liquid crystal display device.

  The query information acquisition unit 230, the redundancy score calculation unit 240, the search time difference calculation unit 250, the clustering unit 260, the heat map generation unit 270, and the display information generation unit 280, for example, are executed by the processor of the query analysis device 200. May be realized by hardware such as LSI, ASIC, FPGA, or may be realized by cooperation of software and hardware.

  The storage unit 290 is realized by, for example, a RAM, ROM, HDD, flash memory, or a hybrid storage device in which a plurality of these are combined. Further, a part or all of the storage unit 290 may be an external device accessible by the query analysis device 200, such as a NAS or an external storage server.

  The user terminal 300 is a terminal used by a user and is a mobile phone such as a smartphone or a tablet terminal, but is not limited thereto. For example, the user terminal 300 may be a notebook computer, a desktop computer, or a PDA.

  The user terminal 300 includes a control unit 310, an input unit 320, and a display unit 330. For example, the control unit 310 may be realized by a processor of the user terminal 300 executing a program, or may be realized by hardware such as an LSI, an ASIC, or an FPGA, or software and hardware cooperate with each other. It may be realized by doing.

  The input unit 320 is an input device such as a touch panel. When the user terminal 300 is a notebook computer or a desktop computer, the input unit 320 may be an input device such as a keyboard or a mouse. The display unit 330 is a display device such as a liquid crystal display device.

<2. Web page search processing>
FIG. 2 is a sequence diagram illustrating web page search processing according to the embodiment. First, the user uses the input unit 320 of the user terminal 300 to input an instruction for displaying the search page on the display unit 330. A search page is a page of a search site provided by the operator of the web server 100. The control unit 310 of the user terminal 300 transmits an HTTP (Hypertext Transfer Protocol) request to the web server 100 based on the input instruction (S10).

  When receiving the HTTP request from the user terminal 300, the control unit 110 of the web server 100 reads the search page generation information stored in the storage unit 120 in advance (S11). Next, the control unit 110 of the web server 100 transmits the read search page generation information to the user terminal 300 (S12). The search page generation information is, for example, text data describing HTML (HyperText Markup Language), style sheets, image data, moving image data, and audio data.

  When receiving the search page generation information from the web server 100, the control unit 310 of the user terminal 300 causes the display unit 330 to display the search page using the received search page generation information (S13). The user uses the input unit 320 of the user terminal 300 to input a query to the search page displayed on the display unit 330. A query is a single search word or a combination of multiple search words entered on the search page.

  The control unit 310 of the user terminal 300 generates query information based on the query input by the user. The query information includes user identification information, search date and time, in addition to the query input by the user. The control unit 310 transmits the generated query information to the web server 100 (S14).

  When receiving the query information from the user terminal 300, the control unit 110 of the web server 100 stores the received query information in the storage unit 120 (S15). Specifically, the control unit 110 classifies the query information for each user based on the user identification information included in the query information received from the user terminal 300 and stores the query information in the storage unit 120.

  Next, the control unit 110 performs a search process based on the query included in the received query information (S16). In the search process, the control unit 110 generates a list of pages including the search word indicated by the query as a search result. Thereafter, the control unit 110 transmits the generated search result to the user terminal 300 (S17).

  When receiving the search result from the web server 100, the control unit 310 of the user terminal 300 displays the received search result on the display unit 330 (S18). Thus, the search process based on the query input by the user is performed.

  In the present embodiment, the web server 100 is provided with a search function for the sake of simplification of explanation, but the present invention is not limited to this. For example, the query analysis system 10 may include a web server and a search server separately.

<3. Duplication score calculation processing>
Next, the calculation process of the redundancy score will be described. The duplication degree score is a value indicating the degree of duplication between a user who searches for a certain query and a user who searches for another query, and is used as an index for query analysis. The multiplicity score calculation unit 240 of the query analysis device 200 calculates a multiplicity score using the query information.

FIG. 3 is a diagram illustrating an example of query information stored in the storage unit 120 according to the embodiment. As shown in FIG. 3, the storage unit 120 stores query information Q1 to Qn (n: total number of users) classified for each user. For example, the query information Q1 is query information of the user 1, the query information Q2 is query information of the user 2, ..., query information Qn
Is the query information of user n. In each of the query information Q1 to Qn, the query input by the user and the search date and time are associated.

  FIG. 4 is a diagram for explaining the calculation process of the redundancy score according to the embodiment. In FIG. 4, ALLuser indicates the total number of users, Auser indicates the number of users who input the query A, and Buser indicates the number of users who input the query B. Here, ALLuser, Auser, and Buser are the number of users, but may be the number of searches. The values of ALLuser, Auser, and Buser are calculated based on the query information.

  The query information acquisition unit 230 of the query analysis device 200 transmits a query information request to the web server 100. When receiving the query information request from the query information acquisition unit 230, the control unit 110 of the web server 100 reads the query information Q1 to Qn (FIG. 3) from the storage unit 120. Thereafter, the control unit 110 transmits the read query information Q1 to Qn to the query analysis device 200.

  The query information acquisition unit 230 acquires the query information Q1 to Qn transmitted from the control unit 110 of the web server 100. Further, the query information acquisition unit 230 stores the acquired query information Q1 to Qn in the storage unit 290. The duplication degree score calculation unit 240 reads the query information Q1 to Qn from the storage unit 290, and based on the read query information Q1 to Qn, the total number of users ALLuser, the number of users A of the query A, and the number of users B of the query B And calculate.

  Further, the redundancy score calculation unit 240 calculates the redundancy score Score (A, B) of the query B with respect to the query A based on the following formula (1). That is, the redundancy score calculation unit 240 searches the query B by a value obtained by dividing the number of users who searched both the query A and the query B (Auser∩Buser) by the number of users who searched the query A (Auser). Based on the value obtained by dividing the number of users (Buser) by the number of all users (ALLuser), the multiplicity score Score (A, B) is calculated.

  The redundancy score Score (A, B) is a value indicating the degree of overlap between the user who searched the query A and the user who searched the query B. It can be said that the greater the redundancy score Score (A, B), the higher the relevance between query A and query B. Conversely, it can be said that the smaller the multiplicity score Score (A, B), the lower the relevance between the query A and the query B.

  For example, it is assumed that the vehicle type 1 is a light vehicle, the vehicle type 2 is a light vehicle, and the vehicle type 3 is a sports car. In this case, when the query A is “vehicle type 1” and the query B is “vehicle type 2”, the multiplicity score Score (A, B) is a large value. On the other hand, when the query A is “vehicle type 1” and the query B is “vehicle type 3”, the multiplicity score Score (A, B) is a small value. This is because the relationship between light vehicles is high, and the relationship between light vehicles and sports cars is low. In this way, by using the redundancy score Score (A, B), for example, it is possible to extract competitive products and analyze product needs.

<4. Search time difference calculation processing>
Next, search time difference calculation processing will be described. The search time difference is a value indicating the difference between the time when a certain query is searched and the time when another query is searched, and is used as an index for query analysis. The search time difference calculation unit 250 of the query analysis device 200 calculates the search time difference using the query information.

  FIG. 5 is a diagram for explaining search time difference calculation processing according to the embodiment. In FIG. 5, the horizontal axis indicates the time when the search is performed, and the vertical axis indicates the number of users who input the query. The query distribution QA indicates the distribution of the number of users of the query A, and the query distribution QB indicates the distribution of the number of users of the query B. Time T1 indicates the median search time of the query distribution QA, and time T2 indicates the median search time of the query distribution QB.

  The search time difference calculation unit 250 derives the time T1 and the time T2 based on the query information Q1 to Qn acquired by the query information acquisition unit 230. Specifically, the search time difference calculation unit 250 aggregates the search date / time of the query A based on the query information Q1 to Qn, and derives the median of the aggregated search date / time as the time T1. In addition, the search time difference calculation unit 250 aggregates the search date and time of the query B based on the query information Q1 to Qn, and derives the median of the aggregated search date and time as time T2.

  Further, the search time difference calculation unit 250 calculates the search time difference D (A, B) between the query distribution QA and the query distribution QB by subtracting the time T1 from the derived time T2. When the search time difference D (A, B) is a positive value, it can be said that the search for the query B is performed more frequently than the search for the query A. When the search time difference D (A, B) is a negative value, it can be said that the frequency of the search for the query B before the search for the query A is high. Further, it can be said that the closer the search time difference D (A, B) is to 0, the higher the frequency of the search for the query B at the same time as the search for the query A.

  For example, when query A is “cough medicine” and query B is “throat pain”, the search time difference is a negative value. This is because sore throat is an early symptom. On the other hand, when query A is “cough medicine” and query B is “pneumonia”, the search time difference is a positive value. This is because pneumonia is a state in which symptoms have progressed. In this way, by using the search time difference, for example, it is possible to grasp a change in the user's needs over time.

<5. Query analysis window>
FIG. 6 is a diagram illustrating an example of the query analysis window W before the analysis start according to the embodiment. The display information generation unit 280 of the query analysis device 200 generates display information for the query analysis window W. The display unit 220 displays the query analysis window W according to the display information generated by the display information generation unit 280. As shown in FIG. 6, the query analysis window W displays a query selection area 221, a data source selection area 222, a threshold value input area 223, a start button 224, and a result display area 225.

  The query selection area 221 is an area for an analyst using the query analysis apparatus 200 to select a query to be analyzed. The data source selection area 222 is an area for an analyst to select a data source of query information. In the example shown in FIG. 6, the query A is selected as the query to be analyzed, and the data source from January 1, 2015 to December 31, 2015 is selected as the data source.

  The threshold value input area 223 functions as an input unit that receives an input of the threshold value of the redundancy score. The start button 224 is a button for the analyst to instruct the start of analysis. The result display area 225 is an area where the analysis result is displayed. In the result display area 225, the query, the number of users, the redundancy score, and the search time difference are displayed as analysis results.

  The search time difference calculation unit 250 does not calculate a search time difference for a query in which the redundancy score for the query A calculated by the redundancy score calculation unit 240 is less than the threshold value input to the threshold value input area 223. In the example shown in FIG. 6, 4 is input as the threshold value. For this reason, a search time difference is not calculated for a query having a multiplicity score of less than 4, and the analysis result is omitted from the result display area 225.

  When the analyst clicks the start button 224 using the input unit 210 of the query analysis device 200, the query analysis process is started. In the query analysis process, the redundancy score calculation unit 240 calculates a redundancy score, and the search time difference calculation unit 250 calculates a search time difference. The calculated redundancy score and search time difference are displayed in the result display area 225.

  FIG. 7 is a diagram illustrating an example of the query analysis window W after the analysis according to the embodiment. In the query analysis process, the multiplicity score calculation unit 240 calculates a multiplicity score using the data source indicated in the data source selection area 222 among the query information acquired by the query information acquisition unit 230. Specifically, the multiplicity score calculation unit 240 calculates the multiplicity score of another query for the query indicated in the query selection area 221 based on the above-described equation (1).

  However, if the calculated redundancy score is less than the threshold value input to the threshold value input area 223, the redundancy score calculation unit 240 excludes the analysis result for the query having the redundancy score less than the threshold value from the result display area 225. To do. As a result, it is possible to exclude the analysis result of the query that is less relevant to the query to be analyzed, and to improve the accuracy of the query analysis.

  In the query analysis process, the search time difference calculation unit 250 calculates a search time difference using the data source indicated in the data source selection area 222 among the query information acquired by the query information acquisition unit 230. Specifically, the search time difference calculation unit 250 calculates a search time difference between the query indicated in the query selection area 221 and another query.

  As shown in FIG. 7, in the result display area 225, the redundancy score calculated by the redundancy score calculation unit 240 and the search time difference calculated by the search time difference calculation unit 250 are associated with each query. Is displayed. Thus, the analyst can easily grasp the redundancy score and the search time difference for each query.

  In FIG. 7, the display unit 220 may display a button for rearranging the number of users, the redundancy score, or the search time difference in ascending order or descending order in the query analysis window W. By rearranging these values, the analyst can more easily grasp the analysis result.

<6. Heat map generation process>
When the query analysis process is completed, a heat map display button 226 is displayed in the query analysis window W. When the analyst clicks the heat map display button 226 using the input unit 210 of the query analysis device 200, the heat map generation unit 270 starts the heat map generation process. In the heat map generation process, the clustering unit 260 of the query analysis device 200 generates a cluster by grouping a plurality of queries.

  FIG. 8 is a diagram for explaining clustering processing according to the embodiment. The clustering unit 260 generates a cluster by grouping a plurality of similar queries based on the query information acquired by the query information acquisition unit 230. In the example illustrated in FIG. 8, the clustering unit 260 groups the query C from the query A to generate the cluster A, groups the query G from the query D to generate the cluster B, and groups the query K from the query H. To generate cluster C.

  For example, the clustering unit 260 may generate a cluster by grouping queries including a predetermined number or more of the same search word, or determine whether or not the search word is similar using a thesaurus dictionary. A cluster may be generated by grouping queries including a predetermined number of words.

  The clustering unit 260 outputs the generated cluster to the heat map generation unit 270. The heat map generation unit 270 generates the heat map M based on the query information acquired by the query information acquisition unit 230 and the cluster input from the clustering unit 260.

  FIG. 9 is a diagram illustrating an example of the heat map M according to the embodiment. The horizontal axis of the heat map M indicates a search time difference, and the vertical axis indicates a cluster. In the example shown in FIG. 9, the unit of the search time difference on the horizontal axis is shown as a day, but is not limited to this. For example, if more detailed analysis is required, the unit of the search time difference on the horizontal axis may be indicated as time. When a longer-term analysis result is required, the unit of the search time difference on the horizontal axis may be indicated as a month.

  The numerical value described in each cell indicates the ratio of the number of users (the number of searches) when the sum of one horizontal row is 1. That is, the heat map M is list information in which the search time difference calculated by the search time difference calculation unit 250 and the number of searches corresponding to the search time difference are associated with each cluster grouped by the clustering unit 260. As described above, the heat map generation unit 270 generates a heat map M that expresses the distribution of the search time difference on the horizontal axis and expresses the value obtained by normalizing the number of users (search number) for each search time difference on the vertical axis.

  Each cell in the heat map is hatched with a color corresponding to the number of users (number of searches). In the example shown in FIG. 9, cells having a higher ratio with respect to the sum of the values in one vertical column are hatched in a darker color, but are not limited thereto. For example, a cell having a higher ratio with respect to the sum of the values in one column may be hatched with a lighter color.

  The display unit 220 of the query analysis device 200 may display the heat map M in the query analysis window W or display the heat map M in a window different from the query analysis window W. Thus, the display unit 220 displays the heat map M, so that the analyst can easily grasp the search time difference and the number of searches for each cluster.

<7. Query analysis processing>
FIG. 10 is a flowchart illustrating query analysis processing according to the embodiment. The processing according to this flowchart is executed by the query analysis device 200.

  First, the display unit 220 displays the query analysis window W shown in FIG. 6 (S20). Next, the query analysis device 200 determines whether or not the start button 224 has been clicked (S21). When it is determined that the start button 224 is clicked, the redundancy score calculation unit 240 calculates a redundancy score for each query based on the query information acquired by the query information acquisition unit 230 (S22). Next, the search time difference calculation unit 250 calculates a search time difference for each query based on the query information acquired by the query information acquisition unit 230 (S23).

  Thereafter, as shown in FIG. 7 described above, the display information generation unit 280 calculates the redundancy score calculated by the redundancy score calculation unit 240 and the search time difference calculated by the search time difference calculation unit 250 for each query. Display information to be displayed in association with. The display unit 220 displays the display information generated by the display information generation unit 280 (S24). Further, the display unit 220 displays a heat map display button 226 in the query analysis window W.

  Next, the query analysis device 200 determines whether or not the heat map display button 226 has been clicked (S25). When it is determined that the heat map display button 226 is clicked, the heat map generation unit 270 generates the heat map M shown in FIG. 9 (S26). Thereafter, the display unit 220 displays the heat map M generated by the heat map generation unit 270 (S27), and ends the processing according to this flowchart.

  As described above, the multiplicity score calculation unit 240, based on the query information, the multiplicity score Score (A, B) indicating the degree of duplication between the user who searched the query A and the user who searched the query B. Is calculated. The search time difference calculation unit 250 calculates a search time difference D (A, B) that is a difference between the time when the query A is searched and the time when the query B is searched based on the query information. The display information generation unit 280 uses the redundancy score Score (A, B) calculated by the redundancy score calculation unit 240 and the search time difference D (A, B) calculated by the search time difference calculation unit 250 as the query A. And information to be displayed in association with the query B. This makes it possible to analyze the needs of users to be searched and changes in user needs over time using queries.

  Note that the query analysis apparatus 200 according to the above embodiment has a computer system therein. Each process of the query analysis apparatus 200 described above is stored in a computer-readable recording medium in the form of a program, and the above-described various processes are performed by the computer reading and executing the program. Here, the computer-readable recording medium refers to a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, and the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program.

  The query analysis device 200 includes both the display unit 220 and the display information generation unit 280, but is not limited thereto. For example, when the query analysis device 200 is provided as a cloud service, the display unit 220 can be omitted. In this case, the display information generation unit 280 may generate display information for displaying the analysis result, and transmit the generated display information in response to a request from the outside.

  Further, although the search time difference calculation unit 250 derives the median value of the search time and calculates the search time difference using the derived median value, the present invention is not limited to this. For example, the search time difference calculation unit 250 may calculate an average value of search times and calculate a search time difference using the calculated average value.

  Moreover, although the heat map production | generation part 270 produced | generated the heat map with which the search time difference and the number of searches were linked | related for every cluster grouped by the clustering part 260, it is not restricted to this. For example, the heat map generation unit 270 may generate a heat map in which a search time difference and the number of searches are associated with each query. Thus, the analyst can check the heat map M for each query.

  As mentioned above, although the form for implementing this invention was demonstrated using embodiment, this invention is not limited to such embodiment at all, In the range which does not deviate from the summary of this invention, various deformation | transformation and substitution Can be added.

DESCRIPTION OF SYMBOLS 10 ... Query analysis system 100 ... Web server 110 ... Control part 120 ... Memory | storage part 200 ... Query analysis apparatus 210 ... Input part 220 ... Display part 230 ... Query information acquisition part 240 ... Duplication degree score calculation part 250 ... Search time difference calculation part 260 ... clustering part 270 ... heat map generation part 280 ... display information generation part 290 ... storage part 300 ... user terminal 310 ... control part 320 ... input part 330 ... display part

Claims (10)

A query information acquisition unit that acquires query information in which a searched query and a time when the query is searched are associated for each user;
Based on the query information acquired by the query information acquisition unit, a redundancy score for calculating a redundancy score indicating a degree of overlap between a user who has searched for the first query and a user who has searched for the second query A calculation unit;
Based on the query information acquired by the query information acquisition unit, a search time difference that calculates a search time difference that is a difference between a time when the first query is searched and a time when the second query is searched A calculation unit;
Information for displaying the redundancy score calculated by the redundancy score calculation unit and the search time difference calculated by the search time difference calculation unit in association with the first query and the second query A display information generation unit for generating
A query analysis apparatus comprising: The multiplicity score calculation unit includes a value obtained by dividing the number of users who have searched both the first query and the second query by the number of users who have searched the first query, and the second query. The query analysis device according to claim 1, wherein the redundancy score is calculated based on a value obtained by dividing the number of users who have searched for the number of users by the number of all users. The search time difference calculation unit calculates, as the search time difference, a difference between a median time when the first query is searched and a median time when the second query is searched. 2. The query analysis device according to 2. An input unit for receiving an input of a threshold value of the multiplicity score;
The search time difference calculation unit does not calculate the search time difference for a second query in which the redundancy score for the first query calculated by the redundancy score calculation unit is less than the threshold. 4. The query analysis device according to any one of items 1 to 3. The list information generation part which produces | generates the list information with which the said search time difference calculated by the said search time difference calculation part and the number of searches corresponding to the said search time difference were linked | related is provided. The query analysis device described. A clustering unit for grouping a plurality of queries to generate a cluster;
The query analysis device according to claim 5, wherein the list information generation unit generates list information in which the search time difference and the number of searches are associated with each cluster grouped by the clustering unit. The query analysis device according to claim 5, wherein the list information generation unit generates list information in which the search time difference and the number of searches are associated with each query. The list information generation unit generates the list information that expresses a distribution of the search time difference on a first axis and expresses a value obtained by normalizing the number of searches for each search time difference on a second axis. The query analysis device according to any one of 1 to 7. A query information acquisition step of acquiring query information in which a searched query and a time when the query is searched are associated for each user;
Based on the query information acquired in the query information acquisition step, a redundancy score for calculating a redundancy score indicating a degree of overlap between the user who has searched for the first query and the user who has searched for the second query A calculation process;
Based on the query information acquired in the query information acquisition step, a search time difference for calculating a search time difference that is a difference between a time when the first query is searched and a time when the second query is searched A calculation process;
Information for displaying the redundancy score calculated in the redundancy score calculation step and the search time difference calculated in the search time difference calculation step in association with the first query and the second query A display information generation step for generating
A query analysis method comprising: Query information obtained by associating each query with a query searched by a computer and a time when the query was searched,
Based on the acquired query information, a redundancy score indicating a degree of overlap between the user who searched the first query and the user who searched the second query is calculated,
Based on the acquired query information, a search time difference that is a difference between a time when the first query is searched and a time when the second query is searched is calculated,
Generating information for displaying the calculated redundancy score and the calculated search time difference in association with the first query and the second query,
program.

Images