Information processing device, information processing method, and information processing program

Abstract

To provide an information processing apparatus, an information processing method, and an information processing program capable of providing a technique for providing highly convenient content according to a search query.SOLUTION: An information processing apparatus includes a determination unit and an extraction unit. The determination unit determines relationships between first search condition and each of second search conditions. The extraction unit extracts, based on levels of the relationships determined by the determination unit, one or more second search conditions from among the multiple second search conditions, as corresponding search conditions which are to be used for searching together with the first search condition.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to an information processing device, an information processing method, and an information processing program. [Background technology]

[0002] Conventional technologies for providing various types of information to users are known. For example, Patent Document 1 discloses a technology for providing content corresponding to search queries. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2017-134682 [Overview of the project] [Problems that the invention aims to solve]

[0004] While conventional technologies can provide users with content corresponding to search queries, it is desirable to provide further technologies that can deliver highly convenient content in response to search queries.

[0005] This application was made in view of the above, and aims to provide an information processing device, an information processing method, and an information processing program that can provide technology for providing highly convenient content in response to search queries. [Means for solving the problem]

[0006] The information processing device according to the present application comprises a determination unit and an extraction unit. The determination unit determines the relationship between a first search condition and each of a plurality of second search conditions. Based on the degree of relationship determined by the determination unit, the extraction unit extracts one or more second search conditions from the plurality of second search conditions to be used together with the first search condition as corresponding search conditions for the search. [Effects of the Invention]

[0007] According to one aspect of the embodiment, an effect of being capable of providing a technique for providing highly convenient content in response to a search query is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] [Figure 1] Figure 1 is a diagram showing an example of information processing according to the embodiment. [Figure 2] Figure 2 is a diagram showing an example of a configuration of an information processing system according to the embodiment. [Figure 3] Figure 3 is a diagram showing an example of a configuration of an information processing apparatus according to the embodiment. [Figure 4] Figure 4 is a diagram showing an example of a user information table stored in a user information storage unit according to the embodiment. [Figure 5] Figure 5 is a diagram showing an example of a search log information table stored in a search log information storage unit according to the embodiment. [Figure 6] Figure 6 is a diagram showing an example of a search theme related information table stored in a search theme related information storage unit according to the embodiment. [Figure 7] Figure 7 is a diagram showing an example of a configuration of a selection unit of the information processing apparatus according to the embodiment. [Figure 8] Figure 8 is a diagram showing an example of a configuration of a determination unit of the information processing apparatus according to the embodiment. [Figure 9] Figure 9 is a diagram showing an example of the importance of each of a plurality of feature amounts in a model calculated by a calculation processing unit in a first determination unit of the information processing apparatus according to the embodiment. [Figure 10] Figure 10 is a diagram showing another example of the importance of each of a plurality of feature amounts in a model calculated by a calculation processing unit in a first determination unit of the information processing apparatus according to the embodiment. [Figure 11] Figure 11 is a diagram showing an example of local spatial statistics calculated by a calculation processing unit of the information processing apparatus according to the embodiment. [Figure 12]Figure 12 shows an example of the configuration of the classification unit of the information processing device according to the embodiment. [Figure 13] Figure 13 shows an example of multiple first search conditions that have been classified into multiple clusters by the classification processing unit in the classification unit of the information processing device according to the embodiment. [Figure 14] Figure 14 shows an example of the configuration of the extraction unit of the information processing device according to the embodiment. [Figure 15] Figure 15 shows an example of the processing results of the correction unit and update unit of the information processing device according to the embodiment. [Figure 16] Figure 16 shows an example of search result content provided to the user by the information processing unit according to the embodiment. [Figure 17] Figure 17 shows another example of search result content provided to the user by the information processing unit according to the embodiment. [Figure 18] Figure 18 shows yet another example of search result content provided to the user by the information processing unit according to the embodiment. [Figure 19] Figure 19 is a flowchart showing an example of information processing by the processing unit of the information processing device according to the embodiment. [Figure 20] Figure 20 is a flowchart showing an example of content provision processing by the processing unit of the information processing device according to the embodiment. [Figure 21] Figure 21 is a flowchart showing an example of the search theme determination process by the processing unit of the information processing device according to the embodiment. [Figure 22] Figure 22 is a flowchart showing an example of the sorting order update process performed by the processing unit of the information processing device according to the embodiment. [Figure 23] Figure 23 is a hardware configuration diagram showing an example of a computer that implements the functions of the information processing device according to the embodiment. [Modes for carrying out the invention]

[0009] The following describes in detail, with reference to the drawings, the embodiments for implementing the information processing apparatus, information processing method, and information processing program according to the present application (hereinafter referred to as "embodiments"). Note that these embodiments do not limit the information processing apparatus, information processing method, and information processing program according to the present application. Furthermore, each embodiment can be appropriately combined as long as the processing content is not inconsistent. Also, the same parts are denoted by the same reference numerals in each of the following embodiments, and redundant descriptions are omitted.

[0010] [1. An example of information processing] First, an example of information processing according to the embodiment will be described using Figure 1. Figure 1 is a diagram showing an example of information processing according to the embodiment, which is executed by the information processing device 1.

[0011] Information processing device 1 provides information services to each user U, providing various types of information. For example, information processing device 1 provides a transaction object information service, which provides transaction object information, which is information about various types of transaction objects. Transaction objects may be goods such as movable or immovable property, but may also be various types of services.

[0012] The information processing device 1 provides search results for transaction targets corresponding to each user U's search query to each user U's terminal device 2 in the information provision service.

[0013] For example, the information processing device 1 can provide each user U with search results for rental properties as a transaction target, or with search results for used cars as a transaction target. In the following, rental properties may be simply referred to as rental properties.

[0014] The search results for the items being traded are not limited to search results for rental properties and used cars, but may also include search results for various home appliances, various food products, cameras, various clothing, shoes, bags, various outdoor goods, various insurance products, various travel packages, hotel accommodations, and other items.

[0015] First, the generation and storage of search theme-related information will be explained with reference to Figure 1(a). As shown in Figure 1(a), the information processing device 1 acquires search query history information, which includes information on multiple past search queries (step S1).

[0016] The information for each search query included in the search query history information includes information on multiple search conditions. These search conditions are, for example, search conditions included in the search query and attributes of user U who sent the search query from terminal device 2. Examples of search conditions included in the search query are search words or specified search conditions.

[0017] Search terms, for example, when searching for rental properties, might be "Minato-ku, Tokyo," "1K," and "pet-friendly," while when searching for used cars, they might be "Manufacturer A," "sunroof," and "4WD," but are not limited to these examples. Search terms are also called search keywords.

[0018] Search criteria can include, for example, when searching for rental properties: "Rent under 100,000 yen," "Building less than 10 years old," "Within a 10-minute walk from the station," and "Floor area 25m²." 2 Examples include "and above," and in used car searches, examples include "mileage under 10,000 km," "seater capacity of 7 or more," and "engine displacement of 2000cc or more," but the search is not limited to these examples.

[0019] Thus, the search criteria include elements of search conditions such as "maximum rent," "maximum building age," "minutes walking from the station," "minimum floor area," "maximum mileage," "minimum passenger capacity," and "minimum engine displacement." Note that the search criteria may be the same items as the search words, or they may be specified as search words themselves.

[0020] User U's attributes include, for example, demographic attributes and psychographic attributes. Demographic attributes are demographic attributes and include multiple attribute items such as age, gender, occupation, place of residence, annual income, and family structure. Psychographic attributes are psychological attributes and include multiple attribute items such as lifestyle, values, and interests.

[0021] Next, the information processing device 1 extracts multiple types of corresponding search conditions based on the search query history information obtained in step S1 (step S2). The corresponding search conditions are search conditions used in the search together with predetermined search conditions (for example, the first search condition described later).

[0022] For example, if the search conditions identified in a new search query transmitted from user U's terminal device 2 are corresponding search conditions, the predetermined search conditions described above can be used as search elements together with the corresponding search conditions. For example, if the predetermined search condition is "pets allowed" and the corresponding search conditions are "3LDK", "not 1K", and "exclusive area 25m²", then the predetermined search conditions are "3LDK", "not 1K", and "exclusive area 25m²". 2 That's all.

[0023] In this case, the information processing device 1 will, for example, use the search criteria "3LDK", "not 1K", and "exclusive area 25m²" as search conditions. 2 "Above" or "Not a 1K & floor area 25m²" 2 For search queries that specify "above," it is possible to perform a search that includes "pets allowed" as an additional search condition, or to present "pets allowed" as an additional search condition to user U.

[0024] Furthermore, for example, when a search query is identified as "pets allowed" as a search condition, the information processing device 1 will respond with "3LDK", "not 1K", and "exclusive area 25m²". 2 "The above," and "Not a 1K & 25m² floor area." 2 You can perform a search that includes one or more of the following additional search conditions: "3LDK", "not 1K", "exclusive area 25m²". 2 "The above," and "Not a 1K & 25m² floor area."2 The system can present the user U with one or more of the above as additional search criteria.

[0025] Multiple types of corresponding search conditions include a first corresponding search condition, a second corresponding search condition, and a third corresponding search condition. The first corresponding search condition is a corresponding search condition extracted by the information processing device 1 using a first extraction method.

[0026] The second correspondence search condition is a correspondence search condition extracted by the information processing device 1 using the second extraction method. The third correspondence search condition is a correspondence search condition extracted by the information processing device 1 using the third extraction method.

[0027] First, the first extraction method for extracting the first corresponding search condition will be explained. The information processing device 1 extracts one or more second search conditions from among the multiple second search conditions as the first corresponding search condition, based on the relationship between the first search condition and each of the multiple second search conditions.

[0028] The information processing device 1 identifies each of two or more search conditions from among the multiple search conditions included in the search query history information extracted in step S1 as the first search condition (step S2-1). The first search condition is, for example, a search condition that can be distinguished by binary values ​​(e.g., presence or absence). For example, in a search for rental properties, the first search condition may be "pets allowed," "detached house," "newly built," etc., and in a search for used cars, it may be "sunroof," "4WD," "campervan," etc.

[0029] Next, the information processing device 1 identifies, for example, a search condition other than the first search condition from among the multiple search conditions included in the search query history information obtained in step S1 as the second search condition (step S2-2).

[0030] For example, suppose the first search condition is "pets allowed". In this case, the information processing device 1 identifies one or more search conditions other than "pets allowed" from among the multiple search conditions included in the search query history information as the second search condition. Alternatively, suppose the first search condition is the attribute of user U of terminal device 2 that sent the search query: "male, single, in his 20s". In this case, the information processing device 1 identifies each search condition included in past search queries as the second search condition.

[0031] Next, the information processing device 1 generates a model for each of the multiple second search conditions using machine learning, based on the search query history information obtained in step S1, with each of the second conditions' information as a feature, and the first search condition as the classification target (steps S2-3).

[0032] For example, the information processing device 1 generates data for each past search query, using information indicating the presence or absence of a first search condition as a label and information indicating the presence or absence of each of the multiple second search conditions as a feature, and performs a process to generate training data containing this data for each of the first search conditions.

[0033] The information processing device 1 then uses training data for each of the first search conditions to generate a model for each of the first search conditions using machine learning. In the training data, information indicating the presence of the first search condition is, for example, a label value of 1, and information indicating the absence of the first search condition is, for example, a label value of 0.

[0034] The model generated by the information processing device 1 is, for example, a binary classifier, which takes information indicating the presence or absence of each of a plurality of second search conditions as input and outputs a score for the first search condition. The score for the first search condition is, for example, in the range of 0 to 1.

[0035] Examples of models include, but are not limited to, decision trees, regression models, and neural networks. Examples of decision trees include GBDTs (Gradient Boosting Decision Trees) such as XGBoost (eXtreme Gradient Boosting), LightGBM (Light Gradient Boosting Machine), and CatBoost (Category Boosting), but are not limited to these.

[0036] Next, the information processing device 1 performs a process for each of the first search conditions to calculate the importance of each of the multiple features in the model generated in step S2-3 as a relationship between the first search condition and each of the multiple second search conditions (step S2-4).

[0037] Then, based on the degree of relationship between the first search condition determined in step S2-4 and each of the multiple second search conditions, the information processing device 1 extracts one or more second search conditions from the multiple second search conditions as the first corresponding search condition for each first search condition (step S2-5).

[0038] For example, the information processing device 1 extracts, for each first search condition, the second search condition whose relationship with the first search condition is greater than or equal to a threshold among a plurality of second search conditions, and uses it as the first corresponding search condition.

[0039] Furthermore, the information processing device 1 can also extract a predetermined number of second search conditions from among a plurality of second search conditions, in order of their relationship to the first search condition, as the first corresponding search conditions for each first search condition.

[0040] Next, a second extraction method for extracting the second corresponding search condition will be described. The information processing device 1 classifies the multiple first search conditions into multiple clusters and extracts one or more second search conditions that contribute to the classification from among the multiple second search conditions as the second corresponding search condition.

[0041] For example, the information processing device 1 classifies multiple first search conditions into multiple clusters based on multiple second search conditions other than multiple first search conditions that are used as options for each other.

[0042] The types of first search criteria in the second extraction method include, for example, in the search for rental properties, multiple search criteria related to region, multiple search criteria related to rental price range, and multiple search criteria related to floor plan, and in the search for used cars, multiple search criteria related to region, multiple search criteria related to manufacturer, and multiple search criteria related to sales price range, but are not limited to these examples.

[0043] Search criteria related to region include, for example, prefectures such as Tokyo, Kanagawa, Kyoto, Osaka, etc. Search criteria related to rental price range include, for example, rental price less than 50,000 yen, rental price between 50,000 yen and 100,000 yen, rental price between 100,000 yen and 150,000 yen, etc.

[0044] Search criteria for floor plans include, for example, 1K, 1DK, 1LDK, 2K, 2LDK, etc. Search criteria for manufacturers include, for example, Manufacturer A, Manufacturer B, Manufacturer C, Manufacturer D, etc. Search criteria for sales price range include, for example, sales price under 500,000 yen, sales price between 500,000 yen and 1,000,000 yen, sales price between 1,000,000 yen and 1,500,000 yen, etc.

[0045] For example, the information processing device 1 identifies each of the multiple first search conditions that are used as options from among the multiple search conditions included in the search query history information extracted in step S1, according to the type of first search condition (step S2-1).

[0046] The first type of search criteria, as mentioned above, includes, for example, in the case of searching for rental properties, search criteria related to area, search criteria related to rental price range, and search criteria related to floor plan, and in the case of searching for used cars, search criteria related to area, search criteria related to manufacturer, and search criteria related to sales price range, but is not limited to these examples.

[0047] For example, the information processing device 1 identifies, as multiple first search conditions related to a region, from among multiple search conditions included in the search query history information, for example, Tokyo, Kanagawa, Kyoto, Osaka, ...

[0048] Furthermore, the information processing device 1 identifies, as multiple first search conditions related to rental price ranges, from among multiple search conditions included in the search query history information, for example, rental price less than 50,000, rental price 50,000 or more and less than 100,000, rental price 100,000 or more and less than 150,000, etc.

[0049] Furthermore, the information processing device 1 identifies, for example, 1K, 1DK, 1LDK, 2K, 2LDK, etc., from among multiple search conditions included in the search query history information as multiple first conditions related to the floor plan.

[0050] Furthermore, the information processing device 1 identifies manufacturers A, B, C, D, etc., from among multiple search conditions included in the search query history information as multiple first conditions related to the manufacturer.

[0051] Furthermore, the information processing device 1 identifies, for example, from among multiple search conditions included in the search query history information, sales prices of 500,000 or less, sales prices of 500,000 or more and 1,000,000 or less, sales prices of 1,000,000 or more and 1,500,000 or less, etc. as multiple first conditions related to the sales price range.

[0052] Next, the information processing device 1 identifies, as second search conditions, for each type of first search condition, from among the multiple search conditions included in the search query history information identified in step S1, other than the multiple first search conditions identified by type in step S2-1 (step S2-2).

[0053] For example, suppose the first search condition is a search condition related to a region. In this case, the information processing device 1 identifies search conditions other than the region-related search condition from past search queries as the second search condition, or identifies each search condition included in past search queries by user U, who has a region-related search condition as an attribute, as the second search condition.

[0054] Next, the information processing device 1 performs a process to generate a model for classifying the multiple first search conditions into multiple clusters based on multiple second search conditions other than the multiple first search conditions used as options for each type of first search condition (step S2-3).

[0055] For example, the information processing device 1 generates multiple models that, based on the multiple second search conditions identified in step S2-2, each determine two first search conditions that are different combinations of each other from among the multiple first search conditions.

[0056] Each of these models is, for example, a binary classification model, such as a decision tree, Bayesian network, support vector machine, or neural network, but is not limited to such examples.

[0057] For example, if the first search condition is a search condition related to a region, the information processing device 1 generates a model to distinguish between Tokyo and Kanagawa prefectures, a model to distinguish between Tokyo and Kyoto prefectures, a model to distinguish between Tokyo and Osaka prefectures, a model to distinguish between Kanagawa prefectures and Kyoto prefectures, and so on.

[0058] Furthermore, if the first search condition is a search condition related to a manufacturer, the information processing device 1 generates models for distinguishing between manufacturer A and manufacturer B, between manufacturer A and manufacturer C, between manufacturer A and manufacturer D, between manufacturer B and manufacturer C, and so on.

[0059] Next, the information processing device 1 performs the process of calculating an evaluation index for each of the multiple models generated in step S2-3, for each type of first search condition (step S2-4). The evaluation index is, for example, AUC (Area Under the Rectangle Curve), and is used as the degree of dissimilarity between two first search conditions that are determined by the model. Note that the evaluation index may be an index obtained from the confusion matrix instead of AUC. For example, the evaluation index may be accuracy, recall, etc.

[0060] Next, the information processing device 1 performs a classification process for each type of first search condition, classifying the multiple first search conditions into multiple clusters based on the evaluation index of each of the multiple models calculated in step S2-4 (step S2-5).

[0061] For example, the information processing device 1 generates map information for each type of first search condition, in which the strength of the similarity or dissimilarity between the first search conditions is replaced with the distance between points on a map, using the evaluation index of each of the multiple models, such as MDS (Multi-Dimensional Scaling).

[0062] The information processing device 1 then performs a process to classify multiple first search conditions into multiple clusters based on map information generated using MDS or the like, for each type of first search condition. For example, the information processing device 1 classifies multiple first search conditions into multiple clusters using the k-means method.

[0063] Furthermore, the information processing device 1 can classify multiple first search conditions into multiple clusters using methods such as spectral clustering, Gaussian Mixture Models (GMM), or neural networks instead of the k-means method, and can also classify multiple first search conditions into multiple clusters using other clustering methods.

[0064] Furthermore, in the example described above, the information processing device 1 uses a model that distinguishes between two first search conditions, but it is not limited to such an example. For example, instead of a model that distinguishes between two first search conditions, the information processing device 1 can also classify multiple first search conditions into multiple clusters by using a model that directly classifies multiple first search conditions into multiple clusters using multiple second search conditions.

[0065] Next, the information processing device 1 performs a process to extract a second search condition, which is a characteristic that separates clusters in the classification process of step S2-5, as a second corresponding search condition, for each type of first search condition (step S2-6).

[0066] For example, the information processing device 1 uses the information of each of the multiple second search conditions as features to generate a model for distinguishing the multiple clusters mentioned above, and uses the generated model to extract the second search conditions that serve as features separating the clusters as the second corresponding search conditions.

[0067] A model for distinguishing between multiple clusters includes, for example, a cluster-specific model that uses information from each of several second search conditions as features to classify whether or not an item belongs to the target cluster. If there are five clusters, the cluster-specific model would be a model for each of the five types of clusters. For example, suppose the five types of clusters are cluster C1, cluster C2, cluster C3, cluster C4, and cluster C5.

[0068] In this case, the cluster C1 model is, for example, a model that classifies whether a cluster belongs to cluster C1 or to other clusters C2-C5, and the cluster C2 model is, for example, a model that classifies whether a cluster belongs to cluster C2 or to other clusters C1, C3-C5.

[0069] The information processing device 1 calculates the importance of each of the multiple second search conditions in the model for each cluster, and based on the calculated importance levels, extracts one or more second search conditions from the multiple second search conditions as the second corresponding search conditions for each cluster (each model).

[0070] For example, the information processing device 1 extracts, for each cluster, the second search conditions from among multiple second search conditions whose importance in the model is above a threshold, as corresponding search condition candidates. Alternatively, the information processing device 1 can also extract, for each cluster, a predetermined number of second search conditions from among multiple second search conditions, ordered by their importance in the model, as corresponding search condition candidates for each first search condition.

[0071] The information processing device 1 can extract candidate corresponding search conditions as second corresponding search conditions. Furthermore, the information processing device 1 can also extract candidate corresponding search conditions representing differences with other clusters as second corresponding search conditions. For example, the information processing device 1 can extract candidate corresponding search conditions that are not extracted in other clusters as second corresponding search conditions.

[0072] Furthermore, the information processing device 1 can also extract a second corresponding search condition based on the importance of the multiple second search conditions in the model described above, which distinguishes between two first search conditions using the information of each of the multiple second search conditions as a feature.

[0073] For example, the information processing device 1 extracts second search conditions that contribute to the classification of the first search conditions as corresponding search condition candidates for each first search condition, based on the importance of multiple second search conditions in the model described above for distinguishing between two first search conditions. Second search conditions that contribute to the classification of the first search conditions are, for example, second search conditions whose importance in the model is above a threshold, or second search conditions that are within a predetermined number in descending order of importance in the model.

[0074] In this case, the information processing device 1 determines common candidate corresponding search conditions that are common across multiple first search conditions for each cluster, and extracts common candidates from the clusters being evaluated that are not found in the common candidates of other clusters as second corresponding search conditions. The information processing device 1 can also extract each of the common candidates as a second corresponding search condition.

[0075] Furthermore, the method for determining the second search condition that distinguishes the clusters is not limited to the example described above. The information processing device 1 can also extract the second search condition that distinguishes the clusters as a second corresponding search condition using a known method other than the one described above.

[0076] Next, a third extraction method for extracting the third corresponding search condition will be described. The information processing device 1 extracts one or more second search conditions from among the multiple second search conditions as the third corresponding search condition, based on the regional characteristics of each of the multiple second search conditions other than the multiple geographical first search conditions.

[0077] First, the information processing device 1 identifies several geographical first search conditions from among the multiple search conditions included in the search query history information extracted in step S1 (step S2-1).

[0078] The first geographical search criterion is defined by administrative divisions such as the eight regional divisions, prefectures, and municipalities. Multiple geographical search criterion items, in the case of the eight regional divisions, include multiple search criteria such as Hokkaido, Tohoku, Kanto, Chubu, Kinki, Chugoku, Shikoku, and Kyushu.

[0079] Furthermore, the first geographical search criteria could be multiple criteria such as prefectures, including Tokyo, Kanagawa, Kyoto, Osaka, Fukuoka, and Hokkaido. Note that the first geographical search criteria are not limited to those defined by administrative boundaries, but may also be defined by any other boundary.

[0080] Next, the information processing device 1 identifies the search conditions other than the first search condition from among the multiple search conditions included in the search query history information identified in step S1 as the second search condition (step S2-2).

[0081] For example, the information processing device 1 may identify search conditions other than geographical search conditions from among multiple search conditions included in past search queries as second search conditions, or it may identify each search condition included in past search queries by user U who has geographical attributes as second search conditions.

[0082] Next, the information processing device 1 determines the regionality of each of the multiple second search conditions based on the multiple geographical first search conditions (step S2-3). For example, if a second search condition is one in which some of the multiple regions indicated by the multiple first search conditions are used together as a first search condition more frequently (proportion) than the frequency (proportion) of the remaining regions, the information processing device 1 determines that some of the regions have regionality.

[0083] For example, in a search for used cars, if the second search condition is "4WD," and the first search condition is used more frequently (proportionally) with the second search condition "4WD" than with other regions, the information processing device 1 will determine that the second search condition "4WD" has regional characteristics in "Hokkaido" and "Tohoku."

[0084] The information processing device 1 can determine the regionality of the second search condition, which is based on the corresponding search frequency (corresponding search ratio), which is the frequency (proportion) at which the first search condition is used in the search, by spatial autocorrelation analysis, instead of determining it by directly comparing the high and low corresponding search frequencies (corresponding search ratios) between regions as described above.

[0085] For example, the information processing device 1 performs a spatial autocorrelation analysis for each of the multiple second search conditions based on multiple geographical first search conditions, and then performs a test (determination) of the regionality of each of the multiple second search conditions based on the results of the spatial autocorrelation analysis.

[0086] Spatial autocorrelation analysis is, for example, an analysis of spatial autocorrelation representing the regional interaction of events, and is performed, for example, by calculating a Local Indicator of Spatial Association (LISA).

[0087] The local spatial statistic is, for example, a local spatial autocorrelation measure, G * statistic (Getis and Ord, 1992). For each of a plurality of second search conditions, the information processing apparatus 1 obtains, for each of the plurality of regions indicated by the plurality of first search conditions, G for each of the plurality of regions for each of the plurality of second search conditions * statistic is calculated. The information processing apparatus 1 calculates G * statistic or G * a region with a larger absolute value of the statistic is determined to be a region with higher regionality.

[0088] In addition, the local spatial statistic may include an H statistic (Getis and Ord, 2012). In this case, the information processing apparatus 1 uses the H statistic and G * the object is classified into four quadrants according to the magnitude relationship of the absolute values of the statistics, and G * among the two quadrants with large absolute values of the statistic, the quadrant with a large H statistic can be excluded as having high heterogeneity.

[0089] Note that the information processing apparatus 1 calculates G * the regionality of each of the plurality of second search conditions can also be determined by using the G statistic instead of or in addition to the statistic. Further, the information processing apparatus 1 calculates G * instead of or in addition to the statistic and the G statistic, etc., the local Moran's I (Anselin, 1995) or the local Geary's C (Anselin, 1995) or the like can be used to determine the regionality of each of the plurality of second search conditions.

[0090] Next, the information processing device 1 extracts from the multiple second search conditions that have regional characteristics in the region indicated by the first search condition, based on the regional characteristics of each of the multiple second search conditions determined in step S2-3, as a third corresponding search condition that corresponds to the first search condition (step S2-4).

[0091] For example, in a used car search, suppose the second search condition "4WD" is regionally specific to the areas indicated by the first search conditions "Hokkaido" and "Tohoku" (Hokkaido and Tohoku). In this case, the information processing device 1 extracts the second search condition "4WD" as a third corresponding search condition corresponding to the first search condition "Hokkaido," and as a third corresponding search condition corresponding to the first search condition "Tohoku."

[0092] The method for extracting corresponding search conditions by the information processing device 1 is not limited to the examples described above. Furthermore, the information processing device 1 can extract corresponding search conditions using one or more extraction methods pre-configured for each search target from among a plurality of extraction methods, including the first extraction method, the second extraction method, and the third extraction method.

[0093] The search targets are not limited to the rental properties and used cars mentioned above, but may also include various home appliances, various food products, cameras, various clothing, shoes, bags, various outdoor goods, various insurance products, various travel packages, hotel accommodations, and other items.

[0094] Next, the information processing device 1 determines the main search condition and search theme based on the first search condition and corresponding search condition extracted in step S2 (step S3). For example, the information processing device 1 determines the main search condition and search theme based on the first search condition and the first corresponding search condition.

[0095] Furthermore, the information processing device 1 determines the main search condition and search theme based on the first search condition and the second corresponding search condition. Also, the information processing device 1 determines the main search condition and search theme based on the first search condition and the third corresponding search condition.

[0096] The information processing device 1 determines one of the first search condition and the corresponding search condition as the main search condition, and the other of the first search condition and the corresponding search condition as the search theme. The search theme is a search condition used in addition to the main search condition.

[0097] As a result, the information processing device 1 can, for example, set a search condition that includes the first search condition as a search theme for a search query in which a corresponding search condition is identified as a search condition, or set a search condition that includes the corresponding search condition as a search theme for a search query in which a first search condition is identified as a search condition.

[0098] For example, the first search criterion is "pets allowed," and the corresponding search criteria are "3LDK," "not 1K," and "exclusive area 25m²." 2 The above is the case. In this case, the information processing device 1 will use "3LDK" or "not 1K & exclusive area 25m" as search conditions. 2 For search queries that specify "above," the search condition including "pets allowed" can be used as the search theme.

[0099] Furthermore, for search queries where "pets allowed" is specified as a search criterion, the information processing device 1 will then process "3LDK" or "not 1K & floor area 25m²" 2 You can use search terms that include "above" as your search theme.

[0100] Furthermore, suppose the first search condition is "parking" or "property with parking," and the corresponding search conditions are "minutes walking from the station," "minimum floor area," and "Tokyo." In this case, the information processing device 1 can use the search conditions including "property with parking" as a search theme for a search query that identifies "a large number of minutes walking from the station & outside of Tokyo" as search conditions.

[0101] Furthermore, for search queries that specify "parking lot" or "property with parking lot" as search conditions, the information processing device 1 can set a search theme that includes the search conditions "long distance from the station & outside of Tokyo".

[0102] Next, the information processing device 1 stores the information including the main search condition information and the search theme information determined in step S3 in its internal storage unit as search theme-related information for each search theme or for each main search condition (step S4).

[0103] Next, referring to Figure 1(b), we will explain the process from receiving a search query to providing search result content, and the process of updating the order of search theme information. As shown in Figure 1(b), the information processing device 1 receives a search query transmitted from the user U's terminal device 2 (step S5). The search query includes, for example, information on one or more search conditions entered or selected by the user U.

[0104] The information processing device 1 determines, based on the search theme-related information stored in step S4, whether or not there is a main search condition corresponding to the search condition identified by the search query received in step S5 (step S6).

[0105] The search criteria identified by the search query received in step S5 are, for example, the search criteria included in the search query received in step S5, or the search criteria corresponding to the attributes of user U who submitted the search query received in step S5. The search criteria corresponding to user U's attributes are user U's attributes or information related to user U's attributes.

[0106] For example, if the search condition included in the search query received in step S5 is "3LDK", the information processing device 1 will include "3LDK" in the search condition specified in the search query. Also, if the attribute of user U who submitted the search query received in step S5 is "Hokkaido" as the residential area or access area, the information processing device 1 will include "Hokkaido" in the search condition specified in the search query.

[0107] Next, the information processing device 1 extracts search themes corresponding to the main search conditions determined in step S6, based on the search theme-related information stored in step S4 (step S7).

[0108] For example, "3LDK" or "Not a 1K & 25m² floor area". 2 The main search condition is "above" and the search theme corresponding to this main search condition is "pets allowed". In this case, the information processing device 1 determines that the main search condition determined in step S6 is "3LDK" or "not 1K & floor area 25m²". 2 If the search criteria are "above," then "pets allowed" will be extracted as the search theme.

[0109] Furthermore, let's assume that the primary search conditions are "long walking distance from the station" and "outside Tokyo," and the search theme corresponding to these primary search conditions is "properties with parking." In this case, if the primary search conditions determined in step S6 are "long walking distance from the station" and "outside Tokyo," the information processing device 1 extracts "properties with parking" as the search theme.

[0110] For example, if there are multiple main search conditions corresponding to the search conditions identified by the search query received in step S5, the information processing device 1 extracts a search theme corresponding to each of the main search conditions.

[0111] Furthermore, if there are multiple main search conditions corresponding to the search conditions identified by the search query received in step S5, the information processing device 1 can extract one or more main search conditions from among these multiple main search conditions based on predetermined rules or randomly. In this case, the information processing device 1 extracts a search theme corresponding to each of the one or more main search conditions extracted from among the multiple main search conditions.

[0112] The predetermined rules can be different for each attribute information of user U, different for each main search condition, different for each OS (Operating System) type of terminal device 2, or different for each search target. Examples of OS types include PC OS, smartphone OS, and tablet OS.

[0113] The predetermined rules include, for example, the number and types of primary search criteria to extract from multiple primary search criteria. The types of primary search criteria include, for example, a first type which is a search criterion included in the search query, and a second type which is an attribute of user U who submitted the search query, but are not limited to such examples.

[0114] Next, the information processing device 1 performs a search using the main search conditions that it determined correspond to the search conditions identified by the search query in step S6, and the search themes extracted in step S7 as search themes corresponding to those main search conditions (step S8).

[0115] Next, the information processing device 1 provides the search results content to user U by transmitting it to user U's terminal device 2, which is content containing one or more search theme information that is the result of one or more searches performed in step S8 (step S9).

[0116] If the search results content includes multiple search theme pieces, these pieces of information are arranged side-by-side within the search results content. For example, if a portion of the search results content is displayed on terminal device 2 and the area of ​​the search results content displayed on terminal device 2 changes due to scrolling, the multiple search theme pieces are arranged side-by-side in the scrolling direction.

[0117] Here, let's assume that "3LDK" is the primary search criterion, and the search themes corresponding to this primary search criterion are "pet-friendly" and "newly built." In this case, if the search criterion specified in the search query is "3LDK," the information processing device 1 performs a search process that includes "3LDK" and "pet-friendly" as search criteria, and a search process that includes "3LDK" and "newly built" as search criteria.

[0118] Then, as shown in Figure 1(b), the information processing device 1 provides the search results content to user U by transmitting it to user U's terminal device 2, which is content that includes search theme information, which is the result of a search using "3LDK" and "pets allowed" as search conditions, and search theme information, which is the result of a search using "3LDK" and "newly built" as search conditions.

[0119] Each search theme information included in the search results content provided to user U contains information on multiple trade targets that have been searched for, and user U can select the desired trade target information from among the information on multiple trade targets by operating terminal device 2.

[0120] When user U selects information on a transaction target from among information on multiple transaction targets, terminal device 2 sends a query to information processing device 1 to request content corresponding to the selected transaction target information selected by user U.

[0121] When the information processing device 1 receives a query from the terminal device 2 to request content corresponding to the information of the selected transaction, it provides the content corresponding to the information of the selected transaction to the user U by sending the content corresponding to the information of the selected transaction to the terminal device 2. The content corresponding to the information of the selected transaction is, for example, the landing page of the selected transaction (for example, a page that includes detailed information of the selected transaction and a GUI (Graphical User Interface) for purchasing the selected transaction).

[0122] Furthermore, in the processing of step S8, the information processing device 1 may, in addition to or instead of performing a search using the main search conditions and the search theme, perform a search using the search conditions included in the search query without using the search theme.

[0123] In this case, in step S9, the information processing device 1 provides user U with search results content, which includes the results of a search using the search conditions included in the search query without using a search theme.

[0124] Furthermore, if the information processing device 1 does not include the search results content of a search using the main search conditions and search themes, it may include the information of each of the one or more search themes extracted in step S7 as search theme information in the search results content. Such search theme information includes information for sending a search query that includes the main search conditions and search themes as search conditions from the terminal device 2 in response to the user U's operation.

[0125] User U can select search theme information included in the search results content, thereby causing terminal device 2 to send a search query containing the main search conditions and search theme corresponding to the selected search theme information to information processing device 1. When information processing device 1 receives a search query from terminal device 2 that includes the main search conditions and search theme as search conditions, it performs a search using the main search conditions and search theme as search conditions and sends the content containing the search results to terminal device 2. This allows user U to confirm the search results related to the desired search theme.

[0126] The information processing device 1 obtains the evaluation value of each of the multiple search theme information (step S10). For example, the information processing device 1 can obtain the evaluation value of each of the multiple search theme information from an external information processing device, or it can obtain the evaluation value of each of the multiple search theme information by calculating the evaluation value of each of the multiple search theme information.

[0127] The evaluation value of search theme information is, for example, the click-through rate for search results using the search theme, and the click-through rate of user U for information on the trading targets included in the search theme information. Such a click-through rate is, for example, the CTR (Click Through Rate).

[0128] Furthermore, the evaluation value for search theme information may be the conversion rate of the transaction target whose information is included in the search results using the search theme. Such a conversion rate is, for example, CVR (Conversion Rate). Note that the evaluation value for search theme information is not limited to the example above, and may also be, for example, the average value of evaluation values ​​entered by user U.

[0129] Next, the information processing device 1 corrects the evaluation value of each of the multiple search theme pieces based on a correction value corresponding to the sorting order of each of the multiple search theme pieces (step S11). For example, if multiple search theme pieces are arranged in the vertical direction, which is the scrolling direction, in the search results content, the sorting order of the search theme pieces will be such that the top position in the search results content has the highest sorting order, and the sorting order will decrease as you move downwards.

[0130] In this case, the correction value based on the sorting rank is, for example, larger the lower the sorting rank. For instance, the correction value based on the sorting rank might be set to a larger value each time the sorting rank drops by one position, or to a larger value each time the sorting rank drops by two or more positions.

[0131] Furthermore, a correction value corresponding to the sorting order of search theme information not displayed on terminal device 2 in the first view may be set to a larger value than the correction value corresponding to the sorting order of search theme information displayed on terminal device 2 in the first view. The first view refers to the area of ​​the search results content that is initially displayed on terminal device 2 without user U scrolling.

[0132] In this case, the correction values ​​corresponding to the sorting order of each of the multiple search theme information displayed on terminal device 2 in the first view may be the same for all of them, or they may be larger for lower sorting orders. Also, the correction values ​​corresponding to the sorting order of each of the multiple search theme information not displayed on terminal device 2 in the first view may be the same for all of them, or they may be larger for lower sorting orders.

[0133] Furthermore, when the search theme is obtained using a first corresponding search condition or a first search condition corresponding to the first corresponding search condition, the information processing device 1 may, for example, use a correction value that decreases as importance increases, instead of using a correction value that corresponds to the sorting rank. The information processing device 1 may also use a correction value that corresponds to a combination of sorting rank and importance to correct the correction value of the search theme information.

[0134] Next, the information processing device 1 updates the sorting order of each of the multiple search theme information in the search results content based on the correction result of the evaluation value in step S11 (step S12).

[0135] For example, in step S12, the information processing device 1 can assign the highest ranking to each of the multiple updated search theme information items based on their corrected evaluation values. Furthermore, the information processing device 1 can limit the change in each ranking to within n positions in a single update process. n is, for example, an integer greater than or equal to 2. In a single update process, for example, a search theme information item with a ranking of 5th position will have its ranking updated within the range of 5-n to 5+n positions.

[0136] Furthermore, the information processing device 1 can update only the sorting order of search theme information that is not displayed on the terminal device 2 in the first view, or it can update only the sorting order of search theme information that is displayed on the terminal device 2 in the first view.

[0137] In this way, the information processing device 1 determines the relationship between the first search condition and each of the multiple second search conditions, and based on the strength of this relationship, extracts one or more second search conditions from the multiple second search conditions to be used together with the first search condition as corresponding search conditions. As a result, the information processing device 1 can use the extracted corresponding search conditions to provide, for example, the search theme information mentioned above to the user U, and can provide technology to provide highly convenient content according to the search query.

[0138] Furthermore, the information processing device 1 classifies the multiple first search conditions into multiple clusters based on multiple second search conditions other than the multiple first search conditions used as options for each other, and extracts one or more second search conditions that contribute to the above classification from among the multiple second search conditions as corresponding search conditions to be used for searching together with the first search conditions. As a result, the information processing device 1 can use the extracted corresponding search conditions to provide, for example, the search theme information mentioned above to the user U, and can provide technology for providing highly convenient content according to the search query.

[0139] Furthermore, the information processing device 1 determines the regional characteristics of each of the multiple second search conditions other than the first search conditions based on the multiple geographical first search conditions, and based on these regional characteristics, extracts one or more second search conditions from the multiple second search conditions to be used together with the first search conditions as corresponding search conditions for the search. As a result, the information processing device 1 can use the extracted corresponding search conditions to provide, for example, the search theme information mentioned above to the user U, and can provide technology for providing highly convenient content according to the search query.

[0140] Furthermore, the information processing device 1 acquires evaluation values ​​for each of the multiple search theme information items displayed side-by-side in the content provided to user U, and corrects the evaluation values ​​of each of the multiple search theme information items based on correction values ​​corresponding to the order of each of the multiple search theme information items. Then, the information processing device 1 updates the order of each of the multiple search theme information items in the content based on the correction results of the evaluation values ​​of each of the multiple search theme information items. As a result, the information processing device 1 can appropriately update the order of each of the multiple search theme information items, and can provide technology to deliver highly convenient content according to search queries.

[0141] The configuration of the information processing system, including the information processing device 1 and terminal device 2 that perform such processing, will be described in detail below.

[0142] [2. Configuration of the Information Processing System] Figure 2 shows an example of the configuration of an information processing system according to the embodiment. As shown in Figure 2, the information processing system 100 according to the embodiment includes an information processing device 1 and a plurality of terminal devices 2.

[0143] Multiple terminal devices 2 are used by different users U. Terminal devices 2 are, for example, notebook PCs, smartphones, tablet PCs, or wearable devices. Wearable devices include, but are not limited to, smart glasses or smartwatches.

[0144] Each of the information processing device 1 and the multiple terminal devices 2 are connected to each other via a network N, either by wire or wireless, enabling communication between them. Note that the information processing system 100 shown in Figure 2 may include multiple information processing devices 1, etc.

[0145] Network N includes, for example, WANs (Wide Area Networks) such as the Internet, and mobile communication networks such as LTE (Long Term Evolution), 4G (4th Generation), and 5G (5th Generation: 5th Generation Mobile Communication System).

[0146] Terminal device 2 can connect to network N via short-range wireless communication such as a mobile communication network, Bluetooth®, or Wi-Fi (Local Area Network), and communicate with information processing device 1.

[0147] [3. Configuration of Information Processing Device 1] Figure 3 shows an example of the configuration of an information processing device 1 according to an embodiment. As shown in Figure 3, the information processing device 1 includes a communication unit 10, a storage unit 11, and a processing unit 12.

[0148] [3.1. Communications Section 10] The communication unit 10 is implemented, for example, by a NIC (Network Interface Card). The communication unit 10 is connected to the network N by wire or wireless connection and transmits and receives information with various other devices. For example, the communication unit 10 transmits and receives information with the terminal device 2 via the network N.

[0149] [3.2. Storage section 11] The storage unit 11 is implemented by, for example, semiconductor memory elements such as RAM (Random Access Memory) and flash memory, or by storage devices such as hard disks and optical discs. The storage unit 11 includes a user information storage unit 20, a search log information storage unit 21, and a search theme-related information storage unit 22.

[0150] [3.2.1. User information storage unit 20] The user information storage unit 20 stores various information about user U. Figure 4 is a diagram showing an example of a user information table stored in the user information storage unit 20 according to this embodiment.

[0151] In the example shown in Figure 4, the user information table stored in the user information storage unit 20 includes information on items such as "User ID (Identifier)", "Attribute Information", "Location Information", and "Setting Information".

[0152] The "User ID" is an identifier that identifies User U, and is information assigned to each User U. The "Attribute Information" is attribute information that indicates the attributes of User U associated with the "User ID". User U's attributes include, for example, demographic attributes and psychographic attributes. Demographic attributes are demographic attributes and include multiple attribute items such as age, gender, occupation, place of residence, annual income, and family structure.

[0153] Psychographic attributes are psychological attributes that include multiple attribute items related to lifestyle, values, interests, etc. For example, each of the multiple attribute items in a psychographic attribute is an object of interest to user U, such as cars, clothes, travel, games, camping, motorcycles, trains, home appliances, or computers.

[0154] "Location information" includes location information indicating the current location of user U, which is associated with the "user ID". Location information is, for example, information identified by the source IP (Internet Protocol) address of the search query sent from terminal device 2, and is detected or acquired by processing unit 12 and stored in user information storage unit 20.

[0155] Furthermore, location information may also be information included in detection information detected by terminal device 2 and transmitted from terminal device 2. In this case, location information is acquired from terminal device 2 by processing unit 12 and stored in user information storage unit 20. Note that location information may also be information included in the "attribute information" described above as attribute information of user U.

[0156] "Configuration information" refers to the configuration information of user U associated with the "User ID". Note that the information stored in the user information storage unit 20 is not limited to the information described above and may include various other information related to user U.

[0157] [3.2.2. Search log information storage unit 21] The search log information storage unit 21 stores information on past search queries. Figure 5 shows an example of a search log information table stored in the search log information storage unit 21 according to this embodiment.

[0158] In the example shown in Figure 5, the search log information table stored in the search log information storage unit 21 includes information on items such as "search query ID," "search condition information," and "user ID." The "search query ID" is an identifier that identifies the search query received by the information processing device 1, and is assigned to each search query.

[0159] "Search condition information" includes information on one or more search conditions included in the search query corresponding to the "search query ID". Search condition information can be either search word information or search specification information. Search word information, for example, includes information indicating a string representing the search condition.

[0160] The information of the search criteria may include identification information for the search criteria or information that represents the search criteria as a string, but is not limited to these, as long as the information allows the information processing device 1 to identify the search criteria.

[0161] The "User ID" is the User ID of User U of terminal device 2 that sent the search query corresponding to the "Search Query ID," and is the same as the User ID stored in the User Information Storage Unit 20. The User Information Table may also include, in place of or in addition to the "User ID," information indicating the attributes of User U of terminal device 2 that sent the search query corresponding to the "Search Query ID."

[0162] [3.2.3. Search Theme Related Information Storage Unit 22] The search theme-related information storage unit 22 stores information on past search queries. Figure 6 shows an example of a search theme-related information table stored in the search theme-related information storage unit 22 according to this embodiment.

[0163] In the example shown in Figure 6, the search theme-related information table stored in the search theme-related information storage unit 22 includes information on items such as "search theme ID," "main search conditions," "search theme," and "evaluation value." The "search theme ID" is an identifier that identifies the search theme and is assigned to each search theme.

[0164] The "main search condition" includes information on the main search condition corresponding to the "search theme ID". The "search theme" includes information on the search theme that corresponds to the "search theme ID" and has been extracted by the processing unit 12. The search theme is used, for example, to search for trading targets, but is not limited to such examples. For example, the search theme may be used to search for articles such as news, video content, or music content, or it may be used to search for other targets.

[0165] The "evaluation value" is information indicating the evaluation value of the search theme information for the search theme corresponding to the "search theme ID". The search theme information for the search theme corresponding to the "search theme ID" is, for example, the result of a search using the search theme corresponding to the "search theme ID", or information about the search theme corresponding to the "search theme ID". The search theme information includes, for example, information that represents the search theme as a string, and information for sending a search query that includes the main search condition and the search theme as search conditions from terminal device 2 in response to user U's operation.

[0166] The evaluation value of search theme information is, for example, the click-through rate (CTR) for search results using the search theme, and the click-through rate of user U for information on the trading targets included in the search theme information. Such a click-through rate is, for example, CTR.

[0167] Furthermore, the evaluation value for search theme information may be the conversion rate of the transaction target whose information is included in the search results using the search theme. Such a conversion rate is, for example, CVR. Note that the evaluation value for search theme information is not limited to the example above, and may also be, for example, the average value of evaluation values ​​entered by user U.

[0168] Furthermore, the evaluation value of the search theme information may be the click-through rate (e.g., CTR) of users U to the search theme information if the search theme information is indeed search theme information, or it may be the conversion rate (e.g., CVR) of the target of the transaction included in the search results information using the search theme indicated by clicks on the search theme information.

[0169] [3.3. Processing Unit 12] The processing unit 12 is a controller, and is realized, for example, by a CPU (Central Processing Unit) or MPU (Micro Processing Unit) executing various programs stored in the memory device inside the terminal device 2 using RAM as the working area.

[0170] The processing unit 12 may be partially or entirely implemented by an integrated circuit, such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

[0171] As shown in Figure 3, the processing unit 12 includes an acquisition unit 30, a reception unit 31, a selection unit 32, a correction unit 33, an update unit 34, and a provision unit 35, and realizes or executes the information processing functions and operations described below. Note that the internal configuration of the processing unit 12 is not limited to the configuration shown in Figure 3, and other configurations are also acceptable as long as they perform the information processing described later.

[0172] [3.3.1. Acquisition part 30] The acquisition unit 30 acquires various information from external information processing devices and terminal devices 2 via the communication unit 10, and stores the acquired information in the storage unit 11.

[0173] For example, the acquisition unit 30 acquires user information, which is information about user U, from an external information processing device or terminal device 2 via the communication unit 10, and adds the acquired user information to the user information table of the user information storage unit 20.

[0174] Furthermore, the acquisition unit 30 acquires various types of information from the storage unit 11. For example, the acquisition unit 30 acquires user information, which is information about user U, from the user information storage unit 20 or the like. The user information acquired by the acquisition unit 30 includes, for example, some or all of at least one of the attribute information, location information, and setting information mentioned above.

[0175] Furthermore, the acquisition unit 30 acquires search query history information, which includes information on multiple past search queries, from the storage unit 11. The information for each search query includes information on multiple search conditions. The search conditions included in the search query are, for example, the search conditions included in the search query and the attributes of user U who sent the search query from terminal device 2. The search conditions included in the search query are, for example, search words or search specification conditions.

[0176] For example, the acquisition unit 30 acquires search condition information and user ID from the search log information storage unit 21, etc. The search condition information includes information on the search conditions included in the search query. For example, the acquisition unit 30 acquires attribute information of user U corresponding to the user ID acquired from the search log information storage unit 21 from the user information storage unit 20, etc.

[0177] The acquisition unit 30 identifies a first search condition and a second search condition to be used for selecting a search theme from among the multiple search conditions included in the search query history information acquired as described above, and acquires the information of the identified first search condition and the information of the second search condition from the search query history information.

[0178] For example, the method for obtaining the information of the first search condition and the information of the second search condition used in the first extraction method described above will be explained. The acquisition unit 30 identifies each of two or more search conditions from among the multiple search conditions included in the search query history information as the first search condition.

[0179] The first search criteria are, for example, search criteria that can be distinguished by two values. For example, in a search for rental properties, the first search criteria might be "pets allowed," "detached house," and "newly built," while in a search for used cars, they might be "sunroof," "4WD," and "campervan."

[0180] Furthermore, being able to distinguish between two values ​​means, for example, that if the search condition is "pets allowed," then "pets allowed" can be distinguished by whether or not it is allowed, and if the search condition is "sunroof," then "sunroof" can be distinguished by whether or not it has a sunroof.

[0181] Furthermore, the acquisition unit 30 identifies, for example, a search condition other than the first search condition from among multiple search conditions included in the search query history information as the second search condition.

[0182] For example, suppose the first search condition is "pets allowed". In this case, the information processing device 1 identifies one or more search conditions other than "pets allowed" from among the multiple search conditions included in the search query history information as the second search condition. Alternatively, suppose the first search condition is the attribute of user U of terminal device 2 that sent the search query: "male, single, in his 20s". In this case, the acquisition unit 30 identifies each search condition included in past search queries as the second search condition.

[0183] Next, we will explain how to obtain the information of the first search conditions and the information of the second search conditions used in the second extraction method described above. The acquisition unit 30 identifies each of the multiple first search conditions that are used as options for each other from among the multiple search conditions included in the search query history information, according to the type of first search condition.

[0184] As mentioned above, the types of first search criteria in the second extraction method include, for example, in the search for rental properties, search criteria related to region, search criteria related to rental price range, and search criteria related to floor plan, and in the search for used cars, search criteria related to region, search criteria related to manufacturer, and search criteria related to sales price range, but are not limited to these examples.

[0185] For example, the acquisition unit 30 identifies, as multiple first search conditions related to a region, from among multiple search conditions included in the search query history information, for example, Tokyo, Kanagawa, Kyoto, Osaka, ...

[0186] Furthermore, the acquisition unit 30 identifies, as multiple first search conditions related to rental price ranges, from among multiple search conditions included in the search query history information, for example, rental price less than 50,000, rental price 50,000 or more and less than 100,000, rental price 100,000 or more and less than 150,000, etc.

[0187] Furthermore, the acquisition unit 30 identifies, for example, 1K, 1DK, 1LDK, 2K, 2LDK, etc., from among multiple search conditions included in the search query history information as multiple first conditions related to the floor plan.

[0188] Furthermore, the acquisition unit 30 identifies manufacturers A, B, C, D, ..., etc., from among multiple search conditions included in the search query history information as multiple first conditions related to manufacturers.

[0189] Furthermore, the acquisition unit 30 identifies, for example, from among multiple search conditions included in the search query history information, a sales price of 500,000 or less, a sales price of 500,000 or more and 1,000,000 or less, a sales price of 1,000,000 or more and 1,500,000 or less, etc. as multiple first conditions related to the sales price range.

[0190] Furthermore, the acquisition unit 30 identifies search conditions other than the first search condition as second search conditions for each type of first search condition from among the multiple search conditions included in the search query history information identified as described above.

[0191] For example, suppose the first search condition is a search condition related to a region. In this case, the acquisition unit 30 identifies search conditions other than the search condition related to a region from past search queries as the second search condition, or identifies each search condition included in past search queries by user U who has the search condition related to a region as an attribute as the second search condition.

[0192] Next, we will explain how to obtain the information of the first search conditions and the information of the second search conditions used in the third extraction method described above. The acquisition unit 30 identifies multiple geographical first search conditions from among the multiple search conditions included in the search query history information.

[0193] The first geographical search criterion is defined by administrative divisions such as the eight regional divisions, prefectures, and municipalities. Multiple geographical search criterion items, in the case of the eight regional divisions, include multiple search criteria such as Hokkaido, Tohoku, Kanto, Chubu, Kinki, Chugoku, Shikoku, and Kyushu.

[0194] Furthermore, the first geographical search criteria could be multiple criteria such as prefectures, including Tokyo, Kanagawa, Kyoto, Osaka, Fukuoka, and Hokkaido. Note that the first geographical search criteria are not limited to those defined by administrative boundaries, but may also be defined by any other boundary.

[0195] Next, the acquisition unit 30 identifies the search conditions other than the first search condition from among the multiple search conditions included in the search query history information identified as described above as the second search condition.

[0196] For example, the acquisition unit 30 may identify search conditions other than geographical search conditions from among multiple search conditions included in past search queries as second search conditions, or it may identify each search condition included in past search queries by user U who has geographical attributes as second search conditions.

[0197] The first and second search conditions may be identified by the selection unit 32 instead of the acquisition unit 30. For example, the determination unit 40 and the classification unit 41 can identify the first and second search conditions instead of the acquisition unit 30. Furthermore, the first and second search conditions may be identified manually by an operator using a terminal device (not shown).

[0198] The acquisition unit 30 acquires, for example, search theme-related information, which is information related to the search theme, from the search theme-related information storage unit 22 or the like. The search theme-related information acquired by the acquisition unit 30 from the search theme-related information storage unit 22 includes, for example, information on the main search conditions, information on the search theme, and at least one of the evaluation values ​​for the search theme information. The evaluation value is, as described above, for example, the click-through rate for the search theme information.

[0199] For example, the acquisition unit 30 acquires evaluation values ​​for each of the multiple search theme information items that are displayed side-by-side in the content provided to the user U (for example, the search results content described above). Each of the multiple search theme information items includes the results of a search using the corresponding search theme from among the multiple search themes, or information about the search theme.

[0200] [3.3.2. Reception Department 31] The reception unit 31 receives various requests and information from external information processing devices and terminal devices 2 via the communication unit 10.

[0201] For example, the reception unit 31 receives a search query from the terminal device 2. The data area of ​​such a search query includes, for example, information on one or more search conditions (e.g., a first search condition, a second search condition, etc.), and the header area of ​​such a search query includes, for example, information to identify user U of the terminal device 2 that sent the search query. The information to identify user U may be, but is not limited to, user U's account, browser cookie, or the IP address of terminal device 2.

[0202] The reception unit 31 identifies the user ID of user U of the terminal device 2 that sent the search query, for example, based on information for identifying user U. The acquisition unit 30 determines the attributes of user U of the terminal device 2 that sent the search query, based on the user ID identified by the reception unit 31 and the user information stored by the storage unit 11. Note that the user ID may be the same as user U's account.

[0203] The reception unit 31, for example, receives search queries from terminal devices that include the first search condition or corresponding search condition described above, or receives search queries from terminal devices 2 of user U that have attributes corresponding to the first search condition or corresponding search condition described above.

[0204] The reception unit 31 receives queries from the terminal device 2 to request content corresponding to the information of the selected transaction target, which is the transaction target selected by user U. The reception unit 31 also receives search queries from the terminal device 2 that include the main search condition and the search theme corresponding to the search theme information selected by user U as search conditions.

[0205] Furthermore, the reception unit 31 receives, for example, a setting request from user U. The setting request includes setting information, which is information related to searching. When the reception unit 31 receives a setting request from user U, it adds the setting information included in the setting request to the user information table in the user information storage unit 20.

[0206] [3.3.3. Selection Department 32] The selection unit 32 selects the main search conditions and search themes by determining them based on the search query history information acquired by the acquisition unit 30. The selection unit 32 then adds the information containing the selected main search conditions and search themes to the search theme-related information table stored in the search theme-related information storage unit 22, which is stored as search theme-related information for each search theme or for each main search condition.

[0207] The selection unit 32 can select main search conditions and search themes using multiple extraction methods, including the first extraction method, the second extraction method, and the third extraction method described above. The selection method for main search conditions and search themes in the selection unit 32 will be described in detail below.

[0208] Figure 7 shows an example of the configuration of the selection unit 32 of the information processing device 1 according to the embodiment. As shown in Figure 7, the selection unit 32 comprises a determination unit 40, a classification unit 41, an extraction unit 42, and a decision unit 43. The determination unit 40, classification unit 41, extraction unit 42, and decision unit 43 will be described below.

[0209] [3.3.3.1. Judgment unit 40] The determination unit 40 determines the relationship between the first search condition and each of the multiple second search conditions, and determines the regionality of each of the multiple second search conditions.

[0210] Figure 8 shows an example of the configuration of the determination unit 40 of the information processing device 1 according to the embodiment. As shown in Figure 8, the determination unit 40 comprises a first determination unit 50 and a second determination unit 51. The first determination unit 50 and the second determination unit 51 will be described below.

[0211] [3.3.3.1.1. First determination unit 50] The first determination unit 50 determines the relationship between the first search condition and each of the multiple second search conditions. For example, the first determination unit 50 determines the relationship between the first search condition and each of the multiple second search conditions for each first search condition based on the search query history information acquired by the acquisition unit 30.

[0212] The relationship between the first and second search criteria is, for example, the relationship in terms of whether they are used in combination in a search. The higher the frequency (proportion) in which the first and second search criteria are used together, the stronger the relationship.

[0213] As shown in Figure 8, the first determination unit 50 includes a generation processing unit 55 that generates a model to be classified, and a calculation processing unit 56 that uses the model generated by the generation processing unit 55 to determine the relationship between the first search condition and each of the multiple second search conditions by calculation. The generation processing unit 55 and the calculation processing unit 56 will be described in more detail below.

[0214] [3.3.3.1.1.1. Generation Processing Unit 55] The generation processing unit 55 generates a model that uses the information of each of the multiple second search conditions as features, based on the search query history information acquired by the acquisition unit 30, and uses the first search condition as the classification target.

[0215] For example, the generation processing unit 55 generates a model for each of the multiple second search conditions, using machine learning, based on the search query history information acquired by the acquisition unit 30, where each of the second conditions' information is used as a feature to classify the first search condition.

[0216] The generation processing unit 55 generates data for each past search query, for example, by using information indicating the presence or absence of a first search condition as a label and information indicating the presence or absence of each of the multiple second search conditions as a feature, thereby generating training data that includes this data for each of the first search conditions.

[0217] The generation processing unit 55 then generates a model for each of the first search conditions using machine learning with the training data for each of the first search conditions. In the training data, information indicating the presence of the first search condition is, for example, a label value of 1, and information indicating the absence of the first search condition is, for example, a label value of 0.

[0218] The model is, for example, a binary classifier, which takes information indicating the presence or absence of each of several second search conditions as input and outputs a score for the first search condition. The score for the first search condition is, for example, in the range of 0 to 1.

[0219] Examples of models include, but are not limited to, decision trees, regression models, and neural networks. Examples of decision trees include, but are not limited to, GBDTs such as XGBoost, LightGBM, and CatBoost.

[0220] [3.3.3.1.1.2. Calculation Processing Unit 56] The calculation processing unit 56 calculates the importance of each of the multiple features in the model generated by the generation processing unit 55 as a relationship between each of the multiple second search conditions for each first search condition (for each model).

[0221] For example, the calculation processing unit 56 performs a process for each of the first search conditions to calculate the importance of each of the multiple features in the model generated by the generation processing unit 55 as a relationship between the first search condition and each of the multiple second search conditions.

[0222] For example, if the model generated by the generation processing unit 55 is GBDT, the calculation processing unit 56 defines the feature importance using the split (e.g., the number of divisions) or the gain (e.g., the amount of error reduction when divided). Note that the feature importance is not limited to the examples above and can be calculated using various known techniques.

[0223] Figure 9 is a diagram showing an example of the importance of each of several feature quantities in a model calculated by the calculation processing unit 56 in the first determination unit 50 of the information processing device 1 according to the embodiment. In the example shown in Figure 9, the importance of the second search condition, which is treated as a feature quantity in the model when the first search condition is "parking lot", is shown.

[0224] In the example shown in Figure 9, "maximum rent" is the most important factor, followed by "minimum floor area," "walking distance from the station," "pets allowed," "separate bathroom and toilet," "Tokyo," "air conditioning," "2nd floor or higher," "age of the building," and "laundry machine space," in descending order of importance.

[0225] Figure 10 is a diagram showing another example of the importance of each of the multiple features in the model calculated by the calculation processing unit 56 in the first determination unit 50 of the information processing device 1 according to the embodiment. In the example shown in Figure 10, the importance of the second search condition in the model is shown, which is treated as a feature of the model when the first search condition is "pets allowed".

[0226] In the example shown in Figure 10, "maximum rent" is the most important factor, followed by "parking," "minutes walking from the station," "minimum floor area," "year built," "3LDK," "1K," "minimum rent," "detached house," and "separate bathroom and toilet," in descending order of importance.

[0227] [3.3.3.1.2. Second determination unit 51] The second determination unit 51 determines the regionality of each of the multiple second search conditions other than the first search conditions, based on the multiple geographical first search conditions. The second determination unit 51 determines the regionality of each of the multiple second search conditions for each type of first search condition, for example, based on the search query history information acquired by the acquisition unit 30.

[0228] For example, the second determination unit 51 determines that some of the regions indicated by the multiple first search conditions have regional characteristics if the frequency (proportion) of using some of the regions as the first search conditions in the second search condition is higher than the frequency (proportion) of using the remaining regions.

[0229] The second determination unit 51 determines, for example, in a used car search, if the second search condition is "4WD" and the first search condition is used more frequently (proportionally) with the second search condition "4WD" than with other regions, then the second search condition "4WD" has regional characteristics in "Hokkaido" and "Tohoku".

[0230] The second determination unit 51 can determine the regionality of the second search condition based on the corresponding search frequency (corresponding search ratio), which is the frequency (proportion) at which the second search condition is used in the search together with the first search condition, by spatial autocorrelation analysis, instead of by directly comparing the high and low corresponding search frequencies (corresponding search ratios) between regions as described above.

[0231] For example, the second determination unit 51 performs a spatial autocorrelation analysis for each of the multiple second search conditions based on the multiple first geographical search conditions, and performs a test (determination) of the regionality of each of the multiple second search conditions based on the results of the spatial autocorrelation analysis.

[0232] Spatial autocorrelation analysis is, for example, an analysis of spatial autocorrelation that represents the regional interaction of events, and is performed, for example, by calculating local spatial statistics. Local spatial statistics are, for example, local spatial autocorrelation measures, G * This is a statistic. The second determination unit 51 determines the local spatial statistics for each of the multiple regions indicated by the multiple first search conditions for each of the multiple second search conditions, and the G for each of the multiple regions for each of the multiple second search conditions. * The statistical quantity is calculated. The second determination unit 51 is G * Statistic or G * Regions with larger absolute values ​​of the statistics are judged to have a higher degree of regional characteristics.

[0233] Furthermore, the local spatial statistic may include the H statistic. In this case, the second determination unit 51 determines the H statistic and G * Based on the relative magnitudes of the absolute values ​​of the statistics, the subjects are classified into four quadrants, G * Of the two quadrants with large absolute values ​​of the statistical measures, the quadrant with a large H-statistic can be excluded as it represents greater heterogeneity.

[0234] Furthermore, the second determination unit 51 is G * The regionality of each of the multiple second search conditions can also be determined by using the G statistic in place of or in addition to the statistical measures. Furthermore, the second determination unit 51 uses the G statistic. *In addition to or instead of statistics and G-statistics, local Moran statistics or local Gary statistics can also be used to determine the regionality of each of the multiple second search conditions.

[0235] As shown in Figure 8, the second determination unit 51 includes a calculation processing unit 57 that calculates local spatial statistics for each of the multiple regions for each of the multiple second search conditions, and a determination processing unit 58 that determines the regionality of each of the multiple second search conditions based on the calculation results of the calculation processing unit 57. The calculation processing unit 57 and the determination processing unit 58 will be described in more detail below.

[0236] [3.3.3.1.2.1.Calculation Processing Unit 57] The calculation processing unit 57 calculates local spatial statistics for each of the multiple regions for each of the multiple second search conditions, based on the search query history information acquired by the acquisition unit 30, for each type of first search condition.

[0237] The local spatial statistics calculated by the calculation processing unit 57 are, as described above, for example, a local spatial autocorrelation measure, G * These are statistics, but H-statistics and G * It may also be a combination with statistics, G * The local Moran statistic or the local Gary statistic may be used instead of or in addition to the statistics and G-statistics.

[0238] Figure 11 shows an example of local spatial statistics calculated by the calculation processing unit 57 of the information processing device 1 according to the embodiment. Figure 11(a) shows the search ratio (search frequency) for each geographical first search condition (prefecture) used in the search together with the second search condition "4WD", and Figure 11(b) shows the local spatial statistics for each geographical first search condition (prefecture) used in the search together with the second search condition "4WD".

[0239] In the local spatial statistics shown in Figure 11(b), the local spatial statistics for Hokkaido and the Tohoku prefectures are larger compared to the search proportion (search frequency) shown in Figure 11(a), clearly showing the difference between these local spatial statistics and those of other prefectures. Therefore, it can be easily determined that among the multiple geographical first search conditions (prefectures) used in searches together with the second search condition "4WD", Hokkaido and the Tohoku prefectures are geographical first search conditions that are used in searches with a high proportion (frequency) of the second search condition "4WD".

[0240] [3.3.3.1.2.2. Determination Processing Unit 58] The determination processing unit 58 determines the regionality of each of the multiple second search conditions based on the local spatial statistics calculated by the calculation processing unit 57. For example, the determination processing unit 58 determines G * Statistic or G * Regions with larger absolute values ​​of the statistics are judged to have a higher degree of regional characteristics.

[0241] The determination processing unit 58 determines, for example, that a region has high regional characteristics if the local spatial statistics are above a threshold or if the local spatial statistics are outside a predetermined range. For example, in the example shown in Figure 11(b), the second search condition "4WD" has regional characteristics in Hokkaido and the Tohoku prefectures because the local spatial statistics for Hokkaido and the Tohoku prefectures are above a threshold.

[0242] Furthermore, the determination processing unit 58 determines that the local spatial statistics calculated by the calculation processing unit 57 are the H statistic and G * If it includes statistics, G * Of the two quadrants with large absolute values ​​of the statistics, the quadrant with a large H-statistic is excluded due to its greater heterogeneity, G * Regions with larger absolute values ​​of the statistics are judged to have a higher degree of regional characteristics.

[0243] Furthermore, if the local spatial statistic calculated by the calculation processing unit 57 is a local Moran statistic or a local Gary statistic, the determination processing unit 58 can determine the regionality of each of the multiple second search conditions based on the local Moran statistic or the local Gary statistic.

[0244] [3.3.3.2. Classification section 41] The classification unit 41 performs a process to classify multiple first search conditions into multiple clusters based on multiple second search conditions other than multiple first search conditions used as options for each type of first search condition. For example, the classification unit 41 performs a process to classify multiple first search conditions into multiple clusters based on search query history information acquired by the acquisition unit 30, for each type of first search condition.

[0245] The first type of search criteria, as mentioned above, includes, for example, in the case of searching for rental properties, search criteria related to area, search criteria related to rental price range, and search criteria related to floor plan, and in the case of searching for used cars, search criteria related to area, search criteria related to manufacturer, and search criteria related to sales price range, but is not limited to these examples.

[0246] Figure 12 shows an example of the configuration of the classification unit 41 of the information processing device 1 according to the embodiment. As shown in Figure 12, the classification unit 41 includes a generation processing unit 60 that generates a plurality of models that each distinguish between two different combinations of first search conditions, a calculation processing unit 61 that calculates an evaluation index for each of the plurality of models, and a classification processing unit 62 that classifies the plurality of first search conditions into a plurality of clusters based on the calculation results of the calculation processing unit 61. The generation processing unit 60, the calculation processing unit 61, and the classification processing unit 62 will be described in more detail below.

[0247] [3.3.3.2.1. Generation Processing Unit 60] Based on the search query history information acquired by the acquisition unit 30, the generation processing unit 60 uses the information of each of the multiple second search conditions as a feature and generates multiple models for each type of first search condition, each of which distinguishes two first search conditions that are different from each other among the multiple first search conditions.

[0248] The model generated by the generation processing unit 60 is, for example, a binary classification model, such as a decision tree, Bayesian network, support vector machine, or neural network, but is not limited to these examples.

[0249] For example, if the first search condition is a search condition related to a region, the generation processing unit 60 generates a model that distinguishes between Tokyo and Kanagawa prefectures, a model that distinguishes between Tokyo and Kyoto prefectures, a model that distinguishes between Tokyo and Osaka prefectures, a model that distinguishes between Kanagawa prefectures and Kyoto prefectures, and so on.

[0250] Furthermore, if the first search condition is a search condition related to a manufacturer, the generation processing unit 60 generates models for distinguishing between manufacturer A and manufacturer B, between manufacturer A and manufacturer C, between manufacturer A and manufacturer D, between manufacturer B and manufacturer C, and so on.

[0251] [3.3.3.2.2. Calculation Processing Unit 61] The calculation processing unit 61 calculates the evaluation index for each of the multiple models generated by the generation processing unit 60 for each type of the first search condition.

[0252] The evaluation metric is, for example, AUC, and is used as the dissimilarity between two first search conditions that the model distinguishes. Alternatively, the evaluation metric may be an index obtained from the confusion matrix, for example, accuracy or recall.

[0253] [3.3.3.2.3. Classification Processing Unit 62] The classification processing unit 62 performs, for each type of first search condition, classification processing that classifies a plurality of first search conditions into a plurality of clusters based on the evaluation index of each of the plurality of models calculated by the calculation processing unit 61.

[0254] For example, the classification processing unit 62 uses MDS or the like to convert the evaluation index of each of the plurality of models into map information by replacing the degree of similarity or dissimilarity between the first search conditions with the distance between points on a map, and generates the map information for each type of the first search condition.

[0255] The classification processing unit 62 performs processing of classifying the plurality of first search conditions into a plurality of clusters for each type of first search condition based on map information generated using MDS or the like. For example, the classification processing unit 62 classifies the plurality of first search conditions into a plurality of clusters by the k-means method.

[0256] Note that the classification processing unit 62 can also classify the plurality of first search conditions into a plurality of clusters by using spectral clustering, Gaussian mixture model (GMM), or a neural network instead of the k-means method, and can also classify the plurality of first search conditions into a plurality of clusters by other clustering methods.

[0257] Furthermore, in the example described above, the classification unit 41 uses a model that discriminates between two first search conditions, but the present invention is not limited to this example. For example, instead of using a model that discriminates between two first search conditions, the classification unit 41 can use a model that directly classifies a plurality of first search conditions into a plurality of clusters using a plurality of second search conditions, to classify the plurality of first search conditions into the plurality of clusters.

[0258] In this case, the generation processing unit 60 generates a model that directly classifies a plurality of first search conditions into a plurality of clusters using a plurality of second search conditions, and the classification processing unit 62 uses the model generated by the generation processing unit 60 to classify the plurality of first search conditions into the plurality of clusters.

[0259] Figure 13 shows an example of multiple first search conditions classified into multiple clusters by the classification processing unit 62 in the classification unit 41 of the information processing device 1 according to the embodiment. In the example shown in Figure 13, it is shown that the classification processing unit 62 has classified the multiple first search conditions into four clusters.

[0260] [3.3.3.3. Extraction part 42] The extraction unit 42 extracts one or more second search conditions from among a plurality of second search conditions to be used as corresponding search conditions for the search together with the first search condition. The corresponding search conditions include, for example, a first corresponding search condition, a second corresponding search condition, and a third corresponding search condition, as described above.

[0261] Figure 14 shows an example of the configuration of the extraction unit 42 of the information processing device 1 according to the embodiment. As shown in Figure 14, the extraction unit 42 includes a first extraction unit 70 that extracts a first corresponding search condition using the first extraction method described above, a second extraction unit 71 that extracts a second corresponding search condition using the second extraction method described above, and a third extraction unit 72 that extracts a third corresponding search condition using the third extraction method described above. The first extraction unit 70, the second extraction unit 71, and the third extraction unit 72 will be described in detail below.

[0262] [3.3.3.3.1. First extraction unit 70] The first extraction unit 70 extracts one or more second search conditions from among a plurality of second search conditions as first corresponding search conditions to be used in the search together with the first search condition, based on the degree of relationship determined by the first determination unit 50.

[0263] For example, the first extraction unit 70 extracts, for each first search condition, a second search condition from among a plurality of second search conditions in which the degree of relationship with the first search condition is equal to or greater than a threshold, and uses this as the first corresponding search condition.

[0264] Furthermore, the first extraction unit 70 can also extract a predetermined number of second search conditions from among a plurality of second search conditions, in order of their relationship to the first search condition, as the first corresponding search conditions for each first search condition.

[0265] Furthermore, the first extraction unit 70 can also extract a predetermined number of second search conditions from among a plurality of second search conditions, in order of decreasing relationship with the first search condition and having a relationship equal to or greater than a threshold, as the first corresponding search conditions for each first search condition.

[0266] [3.3.3.3.2. Second extraction section 71] The second extraction unit 71 extracts one or more second search conditions from among a plurality of second search conditions that contribute to the classification by the classification unit 41, to be used as second corresponding search conditions together with the first search conditions. The second extraction unit 71 performs the process of extracting second corresponding search conditions for each type of first search condition.

[0267] For example, the second extraction unit 71 uses the information of each of the multiple second search conditions as features to generate a model for distinguishing the multiple clusters described above, and uses the generated model to extract the second search conditions that serve as features separating the clusters as second corresponding search conditions.

[0268] The second extraction unit 71 includes a model generation unit 75 that generates a model for distinguishing multiple clusters from information on multiple second search conditions, a calculation processing unit 76 that calculates the importance of each of the multiple features in the model for each cluster, and an extraction processing unit 77 that extracts the second corresponding search conditions for each cluster. The model generation unit 75, the calculation processing unit 76, and the extraction processing unit 77 will be described in more detail below.

[0269] [3.3.3.3.2.1. Model generation unit 75] The model generation unit 75 uses the information from each of the multiple second search conditions as features to generate a model that can distinguish between multiple clusters.

[0270] A model for distinguishing between multiple clusters includes, for example, a cluster-specific model that uses information from each of several second search conditions as features to classify whether or not an item belongs to the target cluster. If there are five clusters, for example, there would be a model for each of the five clusters. Let's say the five clusters are cluster C1, cluster C2, cluster C3, cluster C4, and cluster C5.

[0271] In this case, the cluster C1 model is, for example, a model that classifies whether a cluster belongs to cluster C1 or to other clusters C2-C5, and the cluster C2 model is, for example, a model that classifies whether a cluster belongs to cluster C2 or to other clusters C1, C3-C5.

[0272] Models for classifying multiple clusters include, but are not limited to, decision trees, regression models, and neural networks. Decision trees include, but are not limited to, GBDTs such as XGBoost, LightGBM, and CatBoost.

[0273] [3.3.3.3.2.2.Calculation Processing Unit 76] The calculation processing unit 76 calculates the importance of each of the multiple features in the model generated by the model generation unit 75 for each cluster. For example, if the model generated by the model generation unit 75 includes the cluster-specific models described above, the calculation processing unit 76 calculates the importance of each of the multiple features in the cluster-specific models.

[0274] For example, if the model generated by the generation processing unit 55 is GBDT, the calculation processing unit 76 defines the importance of features using the split (e.g., the number of divisions) or the gain (e.g., the amount of error reduction when divided). Note that the importance of features is not limited to the examples above and can be calculated using various known techniques.

[0275] [3.3.3.3.2.3. Extraction Processing Unit 77] The extraction processing unit 77 extracts, for each cluster, one or more second search conditions from among the plurality of second search conditions as a second corresponding search condition, based on the importance of each of the plurality of feature values determined by the calculation processing unit 76.

[0276] For example, the extraction processing unit 77 extracts, for each cluster, as a corresponding search condition candidate a second search condition whose importance in the model is equal to or higher than a threshold value, from among the plurality of second search conditions. Further, the extraction processing unit 77 can also extract, for each first search condition and for each cluster, a predetermined number of second search conditions in descending order of importance in the model from among the plurality of second search conditions as corresponding search condition candidates.

[0277] Further, the extraction processing unit 77 can also extract, for each cluster, a predetermined number of second search conditions in which the importance in the model is equal to or higher than a threshold value and in descending order of importance from among the plurality of second search conditions as corresponding search condition candidates.

[0278] The extraction processing unit 77 can extract the corresponding search condition candidate as the second corresponding search condition. Further, the extraction processing unit 77 can also extract a corresponding search condition candidate that is different from those of other clusters as the second corresponding search condition. For example, the extraction processing unit 77 extracts a corresponding search condition candidate that is not extracted in other clusters as the second corresponding search condition.

[0279] Further, the extraction processing unit 77 can also extract the second corresponding search condition based on the importance of the plurality of second search conditions in the above-described model that discriminates between two first search conditions using information of each of the plurality of second search conditions as feature values.

[0280] For example, the extraction processing unit 77 extracts, for each first search condition, second search conditions that contribute to the classification of the first search condition as corresponding search condition candidates, based on the importance of multiple second search conditions in the model described above for determining the relationship between two first search conditions. Second search conditions that contribute to the classification of the first search condition are, for example, second search conditions whose importance in the model is greater than or equal to a threshold, or a predetermined number of second search conditions in descending order of importance in the model.

[0281] In this case, the extraction processing unit 77 determines common candidate corresponding search conditions that are common to multiple first search conditions for each cluster, and extracts common candidates from the clusters to be evaluated that are not included in the common candidates of other clusters as second corresponding search conditions.

[0282] Furthermore, the method for determining the second search condition that distinguishes the clusters is not limited to the example described above. The second extraction unit 71 can also extract the second search condition that distinguishes the clusters as the second corresponding search condition using a known method other than the one described above.

[0283] [3.3.3.3.3. Third extraction section 72] The third extraction unit 72 extracts one or more second search conditions from among a plurality of second search conditions as third corresponding search conditions to be used in the search together with the first search condition, based on the regional characteristics determined by the second determination unit 51.

[0284] The third extraction unit 72 extracts, for example, a second search condition that has been determined to be regional by the second determination unit 51, together with a first search condition corresponding to a region that has been determined to be regional by the second determination unit 51, as a third corresponding search condition to be used in the search.

[0285] For example, in a search for used cars, suppose the second search condition "4WD" is regionally specific to the areas indicated by the first search conditions "Hokkaido" and "Tohoku" (Hokkaido and Tohoku). In this case, the third extraction unit 72 extracts the second search condition "4WD" as a third corresponding search condition corresponding to the first search condition "Hokkaido," and as a third corresponding search condition corresponding to the first search condition "Tohoku."

[0286] [3.3.3.4. Decision Section 43] The determination unit 43 determines one or more search themes, etc., based on the first search conditions and the corresponding search conditions extracted by the extraction unit 42. For example, the determination unit 43 determines the main search conditions and search themes based on the first search conditions and the first corresponding search conditions.

[0287] Furthermore, the determination unit 43 determines the main search condition and search theme based on the first search condition and the second corresponding search condition. Also, the determination unit 43 determines the main search condition and search theme based on the first search condition and the third corresponding search condition.

[0288] The determination unit 43 determines that one of the first search condition and the corresponding search condition will be the main search condition, and the other of the first search condition and the corresponding search condition will be the search theme. The search theme is a search condition that is used in addition to the main search condition.

[0289] As a result, for example, the decision unit 43 can set a search condition that includes the first search condition as a search theme for a search query in which a corresponding search condition is identified as a search condition, or set a search condition that includes the corresponding search condition as a search theme for a search query in which a first search condition is identified as a search condition.

[0290] For example, if the first search criterion is "pets allowed," and the corresponding search criteria are "3LDK," "not 1K," and "exclusive area 25m²," 2 The above is the case. In this case, the determination unit 43 will use "3LDK" or "not 1K & exclusive area 25m" as search conditions. 2For search queries that specify "above," the search condition including "pets allowed" can be used as the search theme.

[0291] Furthermore, for search queries where "pets allowed" is specified as a search condition, the decision unit 43 determines whether the property is "3LDK" or "not a 1K & 25m² floor area". 2 You can use search terms that include "above" as your search theme.

[0292] Furthermore, suppose the first search condition is "parking" or "property with parking," and the corresponding search conditions are "minutes walking from the station," "minimum floor area," and "Tokyo." In this case, the decision unit 43 can use the search conditions including "property with parking" as a search theme for a search query that identifies "a large number of minutes walking from the station & outside of Tokyo" as search conditions.

[0293] Furthermore, the decision unit 43 can use the search theme to include search conditions such as "long distance from the station & outside of Tokyo" for search queries that specify "parking lot" or "property with parking lot" as search conditions.

[0294] The determination unit 43 adds the information, including the determined main search condition information and the search theme information, to the search theme-related information table stored in the search theme-related information storage unit 22 as search theme-related information for each search theme or for each main search condition.

[0295] [3.3.4. Correction section 33] The correction unit 33 corrects the evaluation value of each of the multiple search theme information based on a correction value corresponding to the sorting order of each of the multiple search theme information. The correction value corresponding to the sorting order can also be called a position bias.

[0296] For example, if multiple search theme information items are arranged vertically in the scroll direction within the search results content, the ranking of the search theme information will be such that the top-ranked item has the highest ranking, and the ranking decreases as you move downwards.

[0297] In this case, the correction value based on the sorting rank is, for example, larger the lower the sorting rank. For instance, the correction value based on the sorting rank might be set to a larger value each time the sorting rank drops by one position, or to a larger value each time the sorting rank drops by two or more positions.

[0298] Furthermore, a correction value corresponding to the sorting order of search theme information not displayed on terminal device 2 in the first view may be set to a larger value than the correction value corresponding to the sorting order of search theme information displayed on terminal device 2 in the first view. The first view refers to the area of ​​the search results content that is initially displayed on terminal device 2 without user U scrolling.

[0299] In this case, the correction values ​​corresponding to the sorting order of each of the multiple search theme information displayed on terminal device 2 in the first view may be the same for all of them, or they may be larger for lower sorting orders. Also, the correction values ​​corresponding to the sorting order of each of the multiple search theme information not displayed on terminal device 2 in the first view may be the same for all of them, or they may be larger for lower sorting orders.

[0300] Furthermore, when a search theme is obtained using a first corresponding search condition or a first search condition corresponding to the first corresponding search condition, the correction unit 33 can, for example, substitute or add to the correction value according to the sorting rank, make the correction value smaller the higher the importance, as described above. The correction unit 33 can also correct the correction value of the search theme information using a correction value according to a combination of sorting rank and importance. The correction unit 33 can also change the correction value according to the sorting rank for each type of search target, for example.

[0301] [3.3.5. Update section 34] The update unit 34 updates the sorting order of each of the multiple search theme information in the content (for example, search result content) based on the correction result of the evaluation value by the correction unit 33.

[0302] For example, the update unit 34 can assign the highest ranking to each of the multiple search theme information items after updating, based on the corrected evaluation value. The update unit 34 can also limit the change in each ranking to within n positions in a single update process. n is, for example, an integer of 2 or more. In a single update process, for example, a search theme information item with a ranking of 5th position will have its ranking updated within the range of 5-n to 5+n positions.

[0303] Furthermore, the update unit 34 can update only the sorting order of search theme information that is not displayed on the terminal device 2 in the first view, or update only the sorting order of search theme information that is displayed on the terminal device 2 in the first view.

[0304] Figure 15 shows an example of the processing results of the correction unit 33 and the update unit 34 of the information processing device 1 according to the embodiment. In the example shown in Figure 15, before updating by the update unit 34, the search theme information with search theme IDs "T1", "T2", "T3", "T4", and "T5" is sorted in the order "3", "1", "5", "4", and "2".

[0305] Furthermore, in the example shown in Figure 15, the correction values ​​for each sorting order are "1", "1.2", "1.3", "1.4", and "1.5" for sorting orders "1", "2", "3", "4", and "5". In this case, the correction unit 33 multiplies the evaluation values ​​of the search theme information for sorting orders "1", "2", "3", "4", and "5" by "1", "1.2", "1.3", "1.4", and "1.5".

[0306] As a result, the evaluation values ​​of the search theme information with the sort order "1", "2", "3", "4", and "5" are corrected to "0.24", "0.25", "0.26", "0.22", and "0.26". Therefore, the update unit 34 updates the sort order of the search theme information with search theme IDs "T1", "T2", "T3", "T4", and "T5" to "1", "4", "2", "5", and "3".

[0307] [3.3.6.Providing Department 35] When a search query is received by the receiving unit 31, the providing unit 35 provides one or more search theme information corresponding to the search query to the user U of the terminal device 2 that sent the search query by transmitting the search theme information to the terminal device 2 that sent the search query.

[0308] For example, when a search query is received by the reception unit 31, the provision unit 35 provides the search results using the main search conditions and the search theme, or information about the search theme, to the terminal device 2 as search theme information.

[0309] The primary search condition is, for example, one of the first search condition and the corresponding search condition, and the search theme is, for example, the other of the first search condition and the corresponding search condition. The search theme information includes, for example, information that represents the search theme as a string, and information for sending a search query that includes the primary search condition and the search theme as search conditions from the terminal device 2 in response to user U's operation.

[0310] The provisioning unit 35 can provide search result content to user U by transmitting the search result content, which is content containing one or more search theme pieces, to user U's terminal device 2.

[0311] Furthermore, when the provision unit 35 provides user U with search result content, which is content containing multiple search theme information, it provides user U with search result content arranged in a sorting order according to the evaluation value (for example, the sorting order before updating by the update unit 34, and the sorting order after updating by the update unit 34).

[0312] As described above, the search query received by the reception unit 31 includes information on one or more search conditions entered or selected by user U. The provision unit 35 determines, based on the search theme-related information stored in the search theme-related information storage unit 22, whether or not there is a main search condition corresponding to the search conditions specified in the search query received by the reception unit 31.

[0313] The primary search condition corresponding to the search conditions specified in the search query is the search condition specified in the search query if there is only one search condition specified in the search query, and one or more of the multiple search conditions specified in the search query if there are multiple search conditions specified in the search query.

[0314] Furthermore, the search criteria specified in the search query are, for example, the search criteria included in the search query, or the search criteria corresponding to the attributes of user U who submitted the search query. The search criteria corresponding to user U's attribute information are user U's attributes or information related to user U's attributes.

[0315] For example, if the search condition included in the search query is "3LDK", the information processing device 1 will determine that the search condition specified in the search query will include "3LDK". Also, if the attribute of user U who submitted the search query is "Hokkaido" as the residential area or access area, the provisioning unit 35 will determine that the search condition specified in the search query will include "Hokkaido".

[0316] If the provisioning unit 35 determines that there is a main search condition corresponding to the search conditions identified by the search query received by the receiving unit 31, it extracts a search theme corresponding to the main search condition based on the search theme-related information stored in the search theme-related information storage unit 22.

[0317] For example, "3LDK" or "Not a 1K & 25m² floor area". 2 The primary search condition is "above" and the search theme corresponding to this primary search condition is "pets allowed". In this case, the provisioning unit 35 determines that the primary search condition identified by the search query received by the receiving unit 31 is "3LDK" or "not 1K & floor area 25m²". 2 If the search criteria are "above," then "pets allowed" will be extracted as the search theme.

[0318] Furthermore, the primary search conditions are "long walking distance from the station" and "outside Tokyo," and the search theme corresponding to these primary search conditions is "parking" or "property with parking." In this case, if the primary search conditions identified in the search query received by the reception unit 31 are "long walking distance from the station" and "outside Tokyo," the provision unit 35 extracts "parking" or "property with parking" as the search theme.

[0319] For example, if there are multiple primary search conditions corresponding to the search conditions specified in the search query, the provisioning unit 35 extracts the search themes corresponding to each of the primary search conditions.

[0320] Furthermore, if there are multiple primary search conditions corresponding to the search conditions specified in the search query, the supply unit 35 can extract one or more primary search conditions from among these multiple primary search conditions based on predetermined rules or randomly. In this case, the supply unit 35 extracts a search theme corresponding to each of the one or more primary search conditions extracted from among the multiple primary search conditions.

[0321] The predetermined rules can be different for each attribute information of user U, different for each main search condition, different for each OS type of terminal device 2, or different for each search target. Examples of OS types include PC operating systems, smartphone operating systems, and tablet operating systems.

[0322] The predetermined rules include, for example, the number and types of primary search criteria to extract from multiple primary search criteria. The types of primary search criteria include, for example, a first type which is a search criterion included in the search query, and a second type which is an attribute of user U who submitted the search query, but are not limited to such examples.

[0323] When the receiving unit 31 receives a search query, the providing unit 35 uses the primary search conditions specified in the search query and the search themes corresponding to those primary search conditions to perform a search for information on the transaction target.

[0324] For example, if the search target is a used car, the providing unit 35 searches for information on used cars using the main search conditions specified in the search query and the search themes corresponding to those main search conditions. Also, if the search target is a rental property, the providing unit 35 searches for information on rental properties using the main search conditions that it has determined correspond to the search conditions specified in the search query and the search themes corresponding to those main search conditions.

[0325] The provisioning unit 35 can search for information on multiple transaction targets stored in the storage unit 11. In this case, the storage unit 11 functions as a search database that stores information on multiple transaction targets in a searchable manner.

[0326] Furthermore, the providing unit 35 can also obtain search results from an external information processing device (for example, a search server). In this case, the providing unit 35 sends a search request to the external information processing device that includes information indicating the type of search target, information on the main search conditions, and information on the search theme. The providing unit 35 then obtains the search results corresponding to the search request from the external information processing device and provides the user U with the content including the obtained search results as search result content.

[0327] The provisioning unit 35 provides the search results content to user U by transmitting the search results content, which is content containing one or more search theme information that is the result of one or more searches using the main search conditions and search themes, to user U's terminal device 2.

[0328] If the search results content includes multiple search theme information, these multiple search theme information items are arranged side by side within the search results content. For example, if the search results content is scrolled and the corresponding area is displayed on terminal device 2, the multiple search theme information items are arranged side by side in the scrolling direction.

[0329] Here, let's assume that "3LDK" is the primary search condition, and the search themes corresponding to this primary search condition are "pet-friendly" and "newly built." In this case, if the search condition specified in the search query is "3LDK," the service provider 35 performs a search process that includes "3LDK" and "pet-friendly" as search conditions, and a search process that includes "3LDK" and "newly built" as search conditions.

[0330] The provision unit 35 then provides the search results content to user U by transmitting search theme information, which is the result of a search using "3LDK" and "pets allowed" as search criteria, and search theme information, which is the result of a search using "3LDK" and "newly built" as search criteria, to user U's terminal device 2.

[0331] Each search theme information included in the search results content provided to user U contains information on multiple trade targets that have been searched for, and user U can select the desired trade target information from among the information on multiple trade targets by operating terminal device 2.

[0332] When user U selects information on a transaction target from among information on multiple transaction targets, terminal device 2 sends a query to information processing device 1 to request content corresponding to the selected transaction target information selected by user U.

[0333] When the provisioning unit 35 receives a query from the terminal device 2 to request content corresponding to the information of the selected transaction target, it provides the content corresponding to the information of the selected transaction target to the user U by sending the content corresponding to the information of the selected transaction target to the terminal device 2. The content corresponding to the information of the selected transaction target is, for example, a landing page for the selected transaction target (for example, a page that includes detailed information of the selected transaction target and a GUI for purchasing the selected transaction target).

[0334] Furthermore, the service provider 35 can also perform a search using the search conditions included in the search query without using a search theme, in addition to or instead of a search using the main search conditions and search theme. In this case, the service provider 35 provides user U with content including the results of a search using the search conditions included in the search query without using a search theme as search result content.

[0335] Furthermore, if the search results using the main search conditions and search themes are not included in the search results content, the provisioning unit 35 may include information for each of one or more search themes as search theme information in the search results content. Such search theme information includes information for sending a search query that includes the main search conditions and search themes as search conditions from the terminal device 2 in response to user U's operation.

[0336] By selecting search theme information included in the search results content, user U can have terminal device 2 send a search query containing the main search conditions and search theme corresponding to the selected search theme information to information processing device 1.

[0337] When the receiving unit 31 receives a search query that includes the main search condition and the search theme as search conditions, the providing unit 35 performs a search using the main search condition and the search theme as search conditions and transmits the content including the search results to the terminal device 2. This allows user U to check the search results related to the desired search theme.

[0338] Figure 16 shows an example of search result content provided to user U by the information processing device 1 according to this embodiment. The search result content 80 shown in Figure 16 includes a search box 81, transaction target information 82a, 82b, 82c, and search theme information 83a, etc.

[0339] The search box 81 contains information about the search criteria entered or selected by user U. In the example shown in Figure 16, the search criterion included in the search box 81 is "3LDK". Each of the transaction item information 82a, 82b, and 82c is information about the transaction item searched using the search criterion "3LDK" entered or selected by user U.

[0340] By operating terminal device 2 and selecting transaction target information 82a, user U can display content showing detailed information about the rental property "XXX Mansion" (the landing page for the rental property "XXX Mansion") on terminal device 2.

[0341] Similarly, by operating terminal device 2 and selecting transaction target information 82a, user U can display content showing detailed information about the rental property "YYY Mansion" (the landing page for the rental property "YYY Mansion") on terminal device 2.

[0342] Similarly, user U can select transaction information 82a by operating terminal device 2, thereby displaying content showing detailed information about the rental property "ZZZ Corpo" (the landing page for the rental property "ZZZ Corpo") on terminal device 2.

[0343] The search theme information 83a contains information on multiple rental properties obtained through a search process that uses the search condition "3LDK" as the main search condition and the search condition "pets allowed" as the search theme. In the example shown in Figure 16, the search theme information 83a contains information on each of the rental properties V1, V2, V3, V4, V5, and V6.

[0344] By operating terminal device 2, user U can select information about rental property V1, thereby displaying content showing detailed information about rental property V1 (the landing page for rental property V1) on terminal device 2. Similarly, user U can display detailed information about rental properties V3, V4, V5, and V6 on terminal device 2 by operating terminal device 2.

[0345] Figure 17 shows another example of search result content 80 provided to user U by the information processing device 1 provisioning unit 35 according to the embodiment. The search result content 80 shown in Figure 17 includes a search box 81 and search theme information 83a, 83b, 83c, 83d, etc.

[0346] The search box 81 shown in Figure 17 is the same as the search box 81 shown in Figure 16. The search theme information 83a shown in Figure 17 is the same as the search theme information 83a shown in Figure 16.

[0347] Search theme information 83b contains information on multiple rental properties obtained through a search process that uses the search condition "3LDK" as the main search condition and the search condition "new construction" as the search theme. In the example shown in Figure 17, search theme information 83b contains information on each of the rental properties W1, W2, W3, W4, W5, and W6.

[0348] User U can select information about rental property W1 by operating terminal device 2, thereby displaying content showing detailed information about rental property W1 (the landing page for rental property W1) on terminal device 2. Similarly, user U can display detailed information about rental properties W3, W4, W5, and W6 on terminal device 2 by operating terminal device 2.

[0349] Search theme information 83c contains information on multiple rental properties obtained through a search process that uses the search condition "3LDK" as the main search condition and the search condition "highly rated by user reviews" as the search theme. In the example shown in Figure 17, search theme information 83c contains information on each of the rental properties X1, X2, X3, X4, X5, and X6.

[0350] By operating terminal device 2, user U can select information about rental property X1, thereby displaying content showing detailed information about rental property X1 (the landing page for rental property X1) on terminal device 2. Similarly, user U can display detailed information about rental properties X3, X4, X5, and X6 on terminal device 2 by operating terminal device 2.

[0351] Search theme information 83b contains information on multiple rental properties obtained through a search process that uses the search condition "3LDK" as the main search condition and the search theme "high satisfaction with surrounding facilities and transportation." In the example shown in Figure 17, search theme information 83c contains information on each of the rental properties Y1, Y2, Y3, Y4, Y5, and Y6.

[0352] By operating terminal device 2, user U can select information about rental property Y1, thereby displaying content showing detailed information about rental property Y1 (the landing page for rental property Y1) on terminal device 2. Similarly, user U can display detailed information about rental properties Y3, Y4, Y5, and Y6 on terminal device 2 by operating terminal device 2.

[0353] In Figure 17, "Pet-friendly" and "Newly built" are search themes selected by the selection unit 32, for example, and "Highly rated by users" and "High satisfaction with surrounding facilities and transportation" are search themes selected by the person setting the search themes, but the examples are not limited to these.

[0354] Furthermore, in Figure 17, the order of the search theme information 83a, 83b, 83c, and 83d is determined and updated by the update unit 34 based on the respective evaluation values ​​of the search theme information 83a, 83b, 83c, and 83d.

[0355] Figure 18 shows yet another example of search result content 80 provided to user U by the provisioning unit 35 of the information processing device 1 according to the embodiment. The search result content 80 shown in Figure 18 includes a search box 81, transaction target information 82a, 82b, 82c, and search theme information 84a, 84b, 84c, 84d, etc.

[0356] The search box 81 shown in Figure 18 is the same as the search box 81 shown in Figure 16. The transaction target information 82a, 82b, and 82c shown in Figure 18 are the same as the transaction target information 82a, 82b, and 82c shown in Figure 16.

[0357] The search theme information 84a, 84b, 84c, and 84d includes information for sending a search query containing the main search condition and the search theme as search conditions from the terminal device 2 in response to user U's operation.

[0358] For example, the search theme information 84a includes information for sending a search query from terminal device 2 in response to user U's operation, which includes the main search condition "3LDK" and the search theme "pets allowed" as search conditions. By operating terminal device 2 and selecting the search theme information 84a, user U can display content on terminal device 2 that includes the search results using the main search condition "3LDK" and the search theme "pets allowed" as search conditions.

[0359] Furthermore, the search theme information 84b includes information for sending a search query containing the main search condition "3LDK" and the search theme "new construction" from the terminal device 2 in response to user U's operation. By operating the terminal device 2 and selecting the search theme information 84a, user U can display content on the terminal device 2 that includes the search results using the main search condition "3LDK" and the search theme "new construction".

[0360] Furthermore, the search theme information 84c includes information for sending a search query from terminal device 2 in response to user U's operation, which includes the main search condition "3LDK" and the search theme "Highly rated by user reviews" as search conditions. By operating terminal device 2 and selecting the search theme information 84a, user U can display content on terminal device 2 that includes the search results using the main search condition "3LDK" and the search theme "Highly rated by user reviews" as search conditions.

[0361] Furthermore, the search theme information 84d includes information for sending a search query from terminal device 2 in response to user U's operation, which includes the main search condition "3LDK" and the search theme "High satisfaction with surrounding facilities and transportation." By operating terminal device 2 and selecting the search theme information 84a, user U can display content on terminal device 2 that includes the search results using the main search condition "3LDK" and the search theme "High satisfaction with surrounding facilities and transportation."

[0362] [4. Processing Procedure] Next, the procedure for information processing by the processing unit 12 of the information processing device 1 according to the embodiment will be described. Figure 19 is a flowchart showing an example of information processing by the processing unit 12 of the information processing device 1 according to the embodiment.

[0363] As shown in Figure 19, the processing unit 12 of the information processing device 1 determines whether or not there is a search query from the terminal device 2 (step S20). If the processing unit 12 determines that there is a search query (step S20: Yes), it performs content provision processing (step S21). The processing in step S21 is the same as the processing in steps S33 to S35 shown in Figure 20, which will be described in detail later.

[0364] When the processing in step S21 is completed, or when it is determined that there are no search queries (step S20: No), the processing unit 12 determines whether it is time to determine the search theme (step S22). The timing for determining the search theme may be, for example, a timing that occurs at a predetermined interval, a timing specified by the operator, or a timing when the number of new search queries reaches a predetermined number, but is not limited to these examples.

[0365] If the processing unit 12 determines that it is time to determine the search theme (step S22: Yes), it performs the search theme determination process (step S23). The process in step S23 is the process in steps S40 to S45 shown in Figure 21, which will be described in detail later.

[0366] If the processing in step S23 is completed, or if it is determined that it is not yet time to determine the search theme (step S22: No), the processing unit 12 determines whether it is time to update the sorting order (step S24). The sorting order update timing is, for example, a timing that occurs at a predetermined interval, but is not limited to such an example.

[0367] If the processing unit 12 determines that it is time to update the sorting order (step S24: Yes), it performs the sorting order update process (step S25). The process in step S25 is the same as the processes in steps S50 to S52 shown in Figure 22, which will be described in detail later.

[0368] If the processing in step S25 is completed, or if it is determined that it is not time to update the sorting order (step S24: No), the processing unit 12 determines whether it is time to terminate the operation (step S26). The processing unit 12 determines that it is time to terminate the operation, for example, when the power to the information processing device 1 is turned off.

[0369] If the processing unit 12 determines that it is not yet time to terminate the operation (step S26: No), it proceeds to step S20. If it determines that it is time to terminate the operation (step S26: Yes), it terminates the process shown in Figure 19.

[0370] Figure 20 is a flowchart showing an example of content provision processing by the processing unit 12 of the information processing device 1 according to the embodiment. As shown in Figure 20, the processing unit 12 identifies search conditions from the search query (step S30).

[0371] Next, the processing unit 12 determines whether there is a main search condition corresponding to the search condition identified in step S30 (step S31). Then, the processing unit 12 identifies a search theme corresponding to the main search condition (step S32).

[0372] Next, the processing unit 12 performs a search based on the main search conditions corresponding to the search conditions identified in step S30 and the search theme identified in step S32 (step S33). Then, the processing unit 12 generates content that includes the results of the search process in step S33 (step S34). The processing unit 12 provides the content generated in step S34 to the user U (step S35), and the process shown in Figure 20 is terminated.

[0373] Figure 21 is a flowchart showing an example of the search theme determination process by the processing unit 12 of the information processing device 1 according to the embodiment. As shown in Figure 21, the processing unit 12 acquires search query history information from the storage unit 11 (step S40).

[0374] Next, the processing unit 12 extracts a first corresponding search condition from the search query history information obtained in step S40 (step S41). The processing unit 12 also extracts a second corresponding search condition from the search query history information obtained in step S40 (step S42). The processing unit 12 also extracts a third corresponding search condition from the search query history information obtained in step S40 (step S43).

[0375] Next, the processing unit 12 determines the main search condition and search theme based on the first to third corresponding search conditions extracted in steps S41 to S43 (step S44). The processing unit 12 stores the search theme-related information, including the main search condition information and search theme information determined in step S44, in the storage unit 11 (step S45), and terminates the process shown in Figure 21.

[0376] Figure 22 is a flowchart showing an example of the sorting order update process performed by the processing unit 12 of the information processing device 1 according to the embodiment. As shown in Figure 22, the processing unit 12 acquires the evaluation value of each of the multiple search theme information (step S50).

[0377] Next, the processing unit 12 corrects the evaluation value of each of the multiple search theme information based on a correction value corresponding to the sorting order of each of the multiple search theme information (step S51). Then, the processing unit 12 updates the sorting order of each of the multiple search theme information in the content based on the evaluation value correction result (step S52), and terminates the process shown in Figure 22.

[0378] [5. Transformation] In the example described above, the extraction unit 42 extracts the first corresponding search condition, the second corresponding search condition, and the third corresponding search condition. However, the system is not limited to this example. For example, it may also extract one or two of the first, second, and third corresponding search conditions as corresponding search conditions depending on the type of search target.

[0379] Furthermore, the extraction unit 42 can also extract one or two of the first corresponding search condition, the second corresponding search condition, and the third corresponding search condition as corresponding search conditions according to the attributes of user U.

[0380] Furthermore, each of the second determination unit 51, the first extraction unit 70, and the second extraction unit 71 can change the aforementioned threshold and predetermined number according to the user U's attributes or the type of search target.

[0381] [6. Hardware Configuration] The information processing device 1 according to the above embodiment is implemented by a computer 200 having a configuration such as that shown in Figure 23. Figure 23 is a hardware configuration diagram showing an example of a computer 200 that implements the functions of the information processing device 1 according to the embodiment. The computer 200 has a CPU 201, RAM 202, ROM (Read Only Memory) 203, HDD (Hard Disk Drive) 204, communication interface (I / F) 205, input / output interface (I / F) 206, and media interface (I / F) 207.

[0382] The CPU 201 operates based on programs stored in the ROM 203 or HDD 204, controlling various components. The ROM 203 stores the boot program executed by the CPU 201 when the computer 200 starts up, as well as programs that depend on the computer 200's hardware.

[0383] HDD204 stores programs executed by CPU201 and data used by such programs. Communication interface205 receives data from other devices via network N (see Figure 2) and sends it to CPU201, and transmits data generated by CPU201 to other devices via network N.

[0384] The CPU 201 controls output devices such as displays and printers, and input devices such as keyboards or mice, via the input / output interface 206. The CPU 201 acquires data from input devices via the input / output interface 206. The CPU 201 also outputs data it has generated to output devices via the input / output interface 206.

[0385] The media interface 207 reads a program or data stored in the recording medium 208 and provides it to the CPU 201 via the RAM 202. The CPU 201 loads the program from the recording medium 208 onto the RAM 202 via the media interface 207 and executes the loaded program. The recording medium 208 can be, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or PD (Phase Change Rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), tape media, magnetic recording medium, or semiconductor memory.

[0386] For example, when the computer 200 functions as the information processing device 1 according to the embodiment, the CPU 201 of the computer 200 realizes the functions of the processing unit 12 by executing a program loaded on the RAM 202. In addition, the data in the storage unit 11 is stored in the HDD 204. The CPU 201 of the computer 200 reads and executes these programs from the recording medium 208, but as another example, these programs may be obtained from other devices via the network N.

[0387] [7. Other] Furthermore, among the processes described in the above embodiments, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically by known methods. In addition, the processing procedures, specific names, and various data and parameters shown in the above document and drawings can be changed at will unless otherwise specified. For example, the various information shown in each figure is not limited to the information shown.

[0388] Furthermore, the components of each illustrated device are functionally conceptual and do not necessarily need to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown, and all or part of them can be functionally or physically distributed and integrated in any unit according to various loads and usage conditions.

[0389] For example, the information processing device 1 described above may be implemented using a terminal device and a server computer, or using multiple server computers. Furthermore, depending on the function, it may be implemented by calling external platforms via APIs (Application Programming Interfaces) or network computing, allowing for flexible configuration changes.

[0390] Furthermore, the embodiments and modifications described above can be combined as appropriate, provided that the processing content is not inconsistent.

[0391] [8. Effects] As described above, the information processing device 1 according to the embodiment comprises a determination unit 40 and an extraction unit 42. The determination unit 40 determines the relationship between the first search condition and each of the plurality of second search conditions. Based on the degree of relationship determined by the determination unit 40, the extraction unit 42 extracts one or more second search conditions from the plurality of second search conditions to be used together with the first search condition as corresponding search conditions for the search. As a result, the information processing device 1 can use the extracted corresponding search conditions to provide, for example, the search theme information described above to the user U, and can provide technology for providing highly convenient content according to the search query.

[0392] Furthermore, the information processing device 1 includes an acquisition unit 30 that acquires search query history information, which includes information on a first search condition and information on a plurality of second search conditions, each containing information on a plurality of past search queries, each containing at least one of these conditions. The determination unit 40 determines the relationship between the first search condition and each of the plurality of second search conditions based on the search query history information acquired by the acquisition unit 30. As a result, the information processing device 1 can provide technology to deliver highly convenient content in response to search queries.

[0393] Furthermore, the determination unit 40 includes a generation processing unit 55 that generates a model that uses the information of each of the multiple second search conditions as features and classifies the first search condition as the target, based on the search query history information acquired by the acquisition unit 30, and a calculation processing unit 56 that calculates the importance of each of the multiple features in the model generated by the generation processing unit 55 as a relationship between each of the multiple second search conditions. As a result, the information processing device 1 can accurately determine the relationship between the first search condition and each of the multiple second search conditions.

[0394] Furthermore, each of the multiple search conditions, including the first search condition and multiple second search conditions, includes at least one of the following: a search word in which information is contained in the search query transmitted from user U's terminal device 2, a search specification condition in which information is contained in the search query, and an attribute of user U. This enables the information processing device 1 to provide a technology for delivering highly convenient content in response to the search query.

[0395] Furthermore, the information processing device 1 includes a receiving unit 31 that receives search queries including corresponding search conditions from a terminal device 2, and a providing unit 35 that, when a search query is received by the receiving unit 31, provides the terminal device 2 with the search results using the first search conditions and corresponding search conditions, or information on the corresponding search conditions. As a result, the information processing device 1 can provide technology for providing highly convenient content in response to search queries.

[0396] Furthermore, the information processing device 1 includes a receiving unit 31 that receives a search query containing a first search condition from a terminal device 2, and a providing unit 35 that, when a search query is received by the receiving unit 31, provides the terminal device 2 with the search results using the first search condition and the corresponding search condition, or information on the corresponding search condition. As a result, the information processing device 1 can provide technology for providing highly convenient content in response to search queries.

[0397] Furthermore, the information processing device 1 includes a receiving unit 31 that receives search queries from a terminal device 2 having attributes corresponding to the corresponding search conditions, and a providing unit 35 that, when a search query is received by the receiving unit 31, provides the results of a search using the first search conditions and the corresponding search conditions, or information about the first search conditions, to the terminal device 2. As a result, the information processing device 1 can provide technology for providing highly convenient content in response to search queries.

[0398] Furthermore, the information processing device 1 includes a receiving unit 31 that receives search queries from a terminal device 2 having attributes corresponding to the first search conditions, and a providing unit 35 that, when a search query is received by the receiving unit 31, provides the terminal device 2 with the search results using the first search conditions and the corresponding search conditions, or information on the corresponding search conditions. As a result, the information processing device 1 can provide technology for providing highly convenient content in response to search queries.

[0399] Furthermore, the information processing device 1 includes a determination unit 43 that determines one or more search themes based on a first search condition and corresponding search conditions extracted by the extraction unit 42, and a provision unit 35 that provides information on one or more search themes. As a result, the information processing device 1 can provide technology for providing highly convenient content in response to search queries.

[0400] Although embodiments of the present application have been described in detail based on the drawings, these are illustrative examples, and the present invention can be implemented in various other forms, including those described in the disclosure section of the invention, based on the knowledge of those skilled in the art.

[0401] Furthermore, the terms "section, module, unit" mentioned above can be replaced with "means" or "circuit," etc. For example, the acquisition unit can be replaced with acquisition means or acquisition circuit. [Explanation of symbols]

[0402] 1. Information Processing Device 2 Terminal devices 10 Communications Department 11 Storage section 12 Processing Units 20 User information storage unit 21 Search log information storage unit 22 Search Theme Related Information Storage Unit 30 Acquisition Department 31 Reception Department 32 Selection Department 33 Correction Unit 34 Update section 35 Providing Department 40 Judgment section 41 Classification Department 42 Extraction part 43 Decision Section 50 First determination unit 51 Second determination unit 55,60 Generation Processing Unit 56, 57, 61, 76 Calculation Processing Unit 58. Determination Processing Unit 62 Classification Processing Unit 70 First extraction section 71 Second extraction section 72 Third extraction section 75 Model Generation Unit 77 Extraction Processing Unit 100 Information Processing Systems N Network

Claims

1. An acquisition unit that acquires search query history information including information on a plurality of past search queries, each containing information on a first search condition and at least one of a plurality of second search condition information, A determination unit determines the relationship between the first search condition and each of the plurality of second search conditions based on the search query history information acquired by the acquisition unit, The system includes an extraction unit that, based on the degree of relationship determined by the determination unit, extracts one or more second search conditions from the plurality of second search conditions as corresponding search conditions to be used in the search together with the first search condition, The determination unit, Based on the search query history information acquired by the acquisition unit, a generation processing unit generates a model that uses the information of each of the plurality of second search conditions as features and sets the first search condition as the classification target. The model generated by the generation processing unit comprises a calculation processing unit that calculates the importance of each of the multiple features in the model, as the relationship between the first search condition and each of the multiple second search conditions, for each of the first search conditions. An information processing device characterized by the following:

2. Each of the plurality of search conditions, including the first search condition and the plurality of second search conditions, The search query transmitted from the user's terminal device includes at least one of the following: a search word containing information, a search specification condition containing information in the search query, and the user's attributes. The information processing apparatus according to feature 1.

3. A receiving unit that receives the search query including the aforementioned corresponding search conditions from the terminal device, The system includes a receiving unit that, when the receiving unit receives the search query, provides the terminal device with the search results using the first search condition and the corresponding search condition, or information about the corresponding search condition. The information processing apparatus according to feature 2.

4. A receiving unit that receives the search query including the first search condition from the terminal device, The system includes a receiving unit that, when the receiving unit receives the search query, provides the terminal device with the search results using the first search condition and the corresponding search condition, or information about the corresponding search condition. The information processing apparatus according to feature 2.

5. A receiving unit that receives the search query from the terminal device having attributes corresponding to the aforementioned search conditions, The system includes a receiving unit that, when the receiving unit receives the search query, provides the terminal device with the search results using the first search condition and the corresponding search condition, or information about the first search condition. The information processing apparatus according to feature 2.

6. A receiving unit that receives the search query from the terminal device having attributes corresponding to the first search condition, The system includes a receiving unit that, when the receiving unit receives the search query, provides the terminal device with the search results using the first search condition and the corresponding search condition, or information about the corresponding search condition. The information processing apparatus according to feature 2.

7. A determination unit that determines one or more search themes based on the first search condition and the corresponding search condition extracted by the extraction unit, A providing unit that provides information on one or more search themes, comprising The information processing apparatus according to feature 1.

8. A method of information processing performed by a computer, A step of acquiring search query history information, which includes information on multiple past search queries, each containing information on a first search criterion and information on at least one of a plurality of second search criteria, A determination step, based on the search query history information obtained by the acquisition step, determines the relationship between the first search condition and each of the plurality of second search conditions, The extraction step includes, based on the degree of relationship determined by the determination step, extracting one or more second search conditions from the plurality of second search conditions as corresponding search conditions to be used in the search together with the first search condition, The aforementioned determination step is, A generation process that generates a model based on the search query history information obtained by the acquisition process, using the information of each of the plurality of second search conditions as features, and classifying the first search condition as the target of classification. The calculation process includes calculating the importance of each of the multiple features in the model generated by the generation process, as the relationship between the first search condition and each of the multiple second search conditions, for each of the first search conditions. An information processing method characterized by the following:

9. A procedure for obtaining search query history information including information on multiple past search queries, each containing information on a first search condition and at least one of a plurality of second search condition pieces, A determination procedure for determining the relationship between the first search condition and each of the plurality of second search conditions based on the search query history information obtained by the acquisition procedure, Based on the degree of relationship determined by the determination procedure, the computer is instructed to perform an extraction procedure which extracts one or more second search conditions from the plurality of second search conditions as corresponding search conditions to be used in the search together with the first search condition. The aforementioned determination procedure is: A generation process procedure that generates a model in which the first search condition is the target of classification, using the information of each of the plurality of second search conditions as features, based on the search query history information obtained by the acquisition procedure, The calculation procedure includes calculating the importance of each of the multiple features in the model generated by the generation procedure, as the relationship between the first search condition and each of the multiple second search conditions, for each of the first search conditions. An information processing program characterized by the following features.

Images