JP7025578B1 - Information processing equipment, information processing methods, and information processing programs - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide more useful information to a user. An information processing apparatus according to the present application includes a generation unit and an output unit. Based on the search history of a user group consisting of multiple users associated with a predetermined action, the generation unit corresponds to a score indicating the relationship between the user group and the predetermined action for each search query included in the search history. The attached model is generated for each predetermined action and updated at a predetermined timing. The output unit outputs information based on the fluctuation of the score for each search query each time the model is generated by the generation unit. [Selection diagram] FIG. 4

Description

This application relates to an information processing device, an information processing method, and an information processing program.

Conventionally, a technique for estimating a relationship with a predetermined matter for each search query has been known.

Japanese Unexamined Patent Publication No. 2016-177377

However, conventional techniques leave room for providing more useful information to users. For example, conventional techniques only estimate the relationship between a given matter and a search query, and do not respond to changes in the relationship.

The present application has been made in view of the above, and an object of the present application is to provide an information processing apparatus, an information processing method, and an information processing program capable of providing more useful information.

The information processing apparatus according to the present application includes a generation unit and an output unit. Based on the search history of a user group consisting of multiple users associated with a predetermined action, the generation unit corresponds to a score indicating the relationship between the user group and the predetermined action for each search query included in the search history. The attached model is generated for each predetermined action and updated at a predetermined timing. The output unit outputs information based on the fluctuation of the score for each search query each time the model is updated by the generation unit.

According to one of the embodiments, more useful information can be provided to the user.

FIG. 1 is a diagram showing an example of information processing according to an embodiment. FIG. 2 is a diagram showing an example of information processing according to an embodiment. FIG. 3 is a diagram showing an example of information processing according to the embodiment. FIG. 4 is a diagram showing a configuration example of the information processing apparatus according to the embodiment. FIG. 5 is a diagram showing an example of a search history according to an embodiment. FIG. 6 is a diagram showing an example of seed user information according to the embodiment. FIG. 7 is a diagram showing an example of model information according to the embodiment. FIG. 8 is a flowchart showing the flow of update processing by the information processing apparatus according to the embodiment. FIG. 9 is a flowchart showing the flow of output processing by the information processing apparatus according to the embodiment. FIG. 10 is a diagram showing an example of seed user information according to a modified example. FIG. 11 is a hardware configuration diagram showing an example of a computer that realizes the functions of the information processing apparatus according to the embodiment and the modified example.

Hereinafter, the information processing apparatus, the information processing method, and the embodiment for implementing the information processing program (hereinafter referred to as “the embodiment”) according to the present application will be described in detail with reference to the drawings. The information processing apparatus, information processing method, and information processing program according to the present application are not limited by the embodiments described below. Further, the embodiments described below can be appropriately combined as long as the processing contents do not contradict each other. Further, in the embodiments described below, the same parts are designated by the same reference numerals, and duplicate explanations are omitted.

[1. Information processing overview]
An example of information processing according to the embodiment will be described with reference to FIGS. 1 to 3. 1 to 3 are diagrams showing an example of information processing according to an embodiment. In the following description, the seed user refers to each user associated with a predetermined action. The predetermined action corresponds to an action such as a search using a certain search query, a visit to a certain facility or a certain store, or a purchase of a certain product. The seed user can also be rephrased as a user who is highly sensitive to a predetermined behavior (or a user who is sensitive to information about a predetermined behavior). For example, a user whose frequency of executing the action AA exceeds a predetermined threshold value can be regarded as a seed user having a high sensitivity to the action AA. The seed user group is a group composed of seed users extracted in advance.

The information processing apparatus 100 according to the embodiment is a model in which a seed user group (an example of a user group associated with a predetermined action) and a score indicating a relationship between the predetermined action are associated with each search query included in the search history. To generate. FIG. 1 shows an example of generating a model for each seed user group associated with behavior AA to behavior AC.

First, when generating a model of a seed user group associated with an action AA, the information processing apparatus 100 sets a seed user group GP ₁₁ composed of a plurality of seed user SUs associated with the action AA. Further, the information processing apparatus 100 acquires the search history of the seed user group GP ₁₁ .

Next, the information processing apparatus 100 is a model in which a score indicating the relationship between the seed user group GP ₁₁ and the action AA is associated with each search query included in the search history based on the search history of the seed user group GP ₁₁ . Generate M ₁₁ . Specifically, the information processing apparatus 100 counts the number of times the seed user SU has entered (or the number of seed user SUs who have entered the search query) for each search query entered by the seed user SU, and the counted number of times. The higher the number, the higher the score is associated with the search query. For example, the information processing apparatus 100 can calculate a score indicating the relationship between the seed user group GP ₁₁ and the action AA for each search query by the following equation (1). The number of types of search queries included in the following formula (1) may be determined based on, for example, the identity of the character strings. The identity is not limited to the case where the character strings are exactly the same, and character strings having a certain degree of similarity may be counted as the same search query.

Score of each search query = Number of times all seed users have entered the search query ÷
Number of corresponding search query types entered by all seed users ... (1)

In this way, the information processing apparatus 100 generates the model M ₁₁ of the seed user group GP ₁₁ associated with the action AA by associating the calculated score with the search query. Similarly, the information processing apparatus 100 can generate a model M ₂₂ of the seed user group GP ₂₂ associated with the action AB and a model M ₃₃ of the seed user group GP ₃₃ associated with the action AC, respectively.

The operator of the information processing apparatus 100 can determine what kind of search query should be selected as a target in the advertising strategy or the like by referring to the above-mentioned models M ₁₁ to M ₃₃ and the like. For example, in the case of an advertising strategy for behavior AA, it can be determined that there is a possibility that the search query associated with the highest score in model M ₁₁ : "Q ₁₁ " can be selected as a target.

The information processing apparatus 100 may further set a comparison user group composed of a plurality of users other than the seed user SU, and generate the above-mentioned model by using the search history of the set comparison user group. The contrast user is a user randomly selected from other than the seed user. Further, the contrast user group is a group composed of the contrast users extracted in advance.

For example, the information processing apparatus 100 specifies a search query that is not entered by the seed user SU among the search queries entered by the comparison user as a non-input query. Then, the information processing apparatus 100 counts the number of times the specified non-input query is input by the comparison user, and associates the score lower (negative value) with the search query as the number of times counted is larger. For example, the information processing apparatus 100 can calculate a score indicating the relationship between the seed user group GP ₁₁ and the action AA for each search query by the following equation (2).

Score of each search query = -Number of times the corresponding search query is entered by all contrast users ÷
Number of corresponding search query types entered by all seed users ... (2)

Further, the information processing apparatus 100 according to the embodiment is mainly characterized in that the above-mentioned models (M ₁₁ to M ₃₃ , etc.) are updated at predetermined timings and information on fluctuations in the score for each search query is output. .. For example, the information processing apparatus 100 includes a model M ₁₁ of the seed user group GP ₁₁ associated with the action AA, a model M ₂₂ of the seed user group GP ₂₂ associated with the action AB, and a seed user group GP ₃₃ associated with the action AC. The model M ₃₃ is updated at a predetermined timing TP ₁ . Then, the information processing apparatus 100 according to the embodiment outputs a search query in which the fluctuation of the score satisfies a predetermined condition.

Thereby, the information processing apparatus 100 can provide the user such as the operator of the information processing apparatus 100 with information regarding the change in the relationship between the search query and the predetermined action. That is, the operator can trace the change in the context of the seed user SU associated with a predetermined action through the change in the score value for each search query output from the information processing apparatus 100, and can detect a peculiar search query that can be a sprout. Is possible. Hereinafter, the main features of the information processing apparatus 100 will be described.

(Pattern 1)
FIG. 2 illustrates an example of a large short-term change in the score of a search query. In FIG. 2, a case where the model M ₁₁ of the seed user group GP ₁₁ associated with the action AA is generated and then the first update timing is reached will be described as an example. The model M ₂₂ of the seed user group GP ₂₂ associated with the action AB and the model M ₃₃ of the seed user group GP ₃₃ associated with the action AC can also be updated by the same method as the example described below.

As shown in FIG. 2, when the information processing apparatus 100 reaches the first update timing TP ₁ (1) predetermined for the seed user group GP ₁₁ , the seed associated with the action AA is associated with the above method. The model M ₁₁ (1) of the user group GP ₁₁ is updated by being generated again (step S2-1). That is, the information processing apparatus 100 is a model M in which a score indicating the relationship between the seed user group GP ₁₁ and the action AA is associated with each search query included in the search history based on the search history of the seed user group GP ₁₁ . ₁₁ (1) is generated again.

Further, the information processing apparatus 100 calculates the variation of the score for each search query of the model M ₁₁ (1), and attempts to output the information regarding the search query in which the variation of the calculated score satisfies a predetermined condition (step S2-2). ). Specifically, the information processing apparatus 100 calculates an increase rate or a decrease rate of the score for each search query between the updated model M ₁₁ (1) and the first model M ₁₁ (0), respectively. Then, the information processing apparatus 100 identifies a search query in which the rate of increase or decrease of the score satisfies a predetermined condition by using the threshold value α and the threshold value β predetermined for determining the fluctuation of the score. For example, the information processing apparatus 100 is a search query that satisfies a predetermined condition when the rate of increase in the score is equal to or higher than a predetermined threshold value α or when the rate of decrease in the score is equal to or less than a predetermined threshold value β. Judge that. In the example shown in FIG. 2, the information processing apparatus 100 outputs a search query: " _Q1N " as a search query in which the increase rate satisfies a predetermined condition, and a search query: "" as a search query in which the decrease rate satisfies a predetermined condition. _Q12 "is being output. When a new search query that was not included in the model before the update appears due to the update of the model, the information processing apparatus 100 calculates the rate of increase or decrease of the score with the predetermined reference value. You may. As the predetermined reference value, the median value or the average value of the scores of all the search queries included in the corresponding model can be adopted.

In this way, the information processing apparatus 100 outputs a search query in which the change in the score value is large each time it is updated. As a result, the operator of the information processing apparatus 100 causes a short-term occurrence in the context of each seed user SU associated with a predetermined action through a change in the relationship between the search query included in the model and the predetermined action corresponding to the model. Can efficiently discover various changes.

(Pattern 2)
FIG. 3 describes an example in which the score of a search query changes gradually over time. In FIG. 3, a case where the model M ₁₁ of the seed user group GP ₁₁ associated with the action AA reaches a certain update timing will be described as an example.

As shown in FIG. 3, when the information processing apparatus 100 reaches a certain update timing TP ₁ (n + n), the model M ₁₁ (n + n) relating to the behavior AA is generated again by the above-mentioned method (step S3-1). That is, the information processing apparatus 100 is a model M in which a score indicating the relationship between the seed user group GP ₁₁ and the action AA is associated with each search query included in the search history based on the search history of the seed user group GP ₁₁ . ₁₁ (n + n) is generated again.

Further, the information processing apparatus 100 calculates the variation of the score for each search query of the model M ₁₁ (n + n) in the same manner as the above-mentioned pattern 1, and the variation of the calculated score satisfies the predetermined condition of the information regarding the search query. Attempt output (step S3-2). Specifically, the information processing apparatus 100 is between the model M ₁₁ (n + n) generated by the update timing TP ₁ (n + n) and the model M ₁₁ (n + m) generated by the immediately preceding update timing TP ₁ (n + m). Calculates the variation in score between each search query. Then, the information processing apparatus 100 identifies a search query in which the rate of increase or decrease of the score satisfies a predetermined condition by using the threshold value α and the threshold value β predetermined for determining the fluctuation of the score. For example, the information processing apparatus 100 is a search query that satisfies a predetermined condition when the rate of increase in the score is equal to or higher than a predetermined threshold value α or when the rate of decrease in the score is equal to or less than a predetermined threshold value β. Judge that. As a result of the process of step S3-2, the information processing apparatus 100 outputs "not applicable" indicating that there is no search query in which the rate of increase or decrease in the score of the search query satisfies a predetermined condition. That is, in the information processing apparatus 100, there is no search query in which there is a change to be output in the score value between the immediately preceding update timing TP ₁ (n + m) and the current update timing TP ₁ (n + n). to decide.

Further, the information processing apparatus 100 calculates the fluctuation of the score for each search query between the model M ₁₁ (n + n) generated by the update timing TP ₁ (n + n) and the first model M ₁₁ (0). An attempt is made to output a search query in which the calculated fluctuation of the score satisfies a predetermined condition (step S3-3).

Specifically, the information processing apparatus 100 increases the rate of increase in the score for each search query between the model M ₁₁ (n + n) generated by the update timing TP ₁ (n + n) and the first model M ₁₁ (0). Or calculate the rate of decline respectively. Then, the information processing apparatus 100 specifies a search query in which the calculated increase rate or decrease rate of the score satisfies a predetermined condition. For example, the information processing apparatus 100 is a search query that satisfies a predetermined condition when the rate of increase in the score is equal to or higher than a predetermined threshold value α or when the rate of decrease in the score is equal to or less than a predetermined threshold value β. To judge. As a result of the process of step S3-3, the information processing apparatus 100 outputs, for example, a search query: "Q ₁₁ " as a search query satisfying a predetermined condition. That is, as described above, the information processing apparatus 100 does not have a search query that causes a remarkable change in the score between the immediately preceding update timing TP ₁ (n + m) and the current update timing TP ₁ (n + n). However, on the other hand, the information processing apparatus 100 generates the first model M ₁₁ and then reaches the update timing TP ₁ (n + n) this time, the search query: "Q ₁₁ ". It is judged that there is a change in the value to be output for the score of.

Further, the information processing apparatus 100 is an example of specifying a search query (a search query exceeding a predetermined threshold value) in which the rate of increase or decrease in the score calculated for each search query of the model M satisfies a predetermined condition. explained. The information processing apparatus 100 is not limited to this example, and may specify a search query showing a peculiar change based on a relative change amount between the search queries of the score associated with the search query.

For example, in the information processing apparatus 100, the score associated with each of the search queries of the model M ₁₁ (0): "Q ₁₁ ", "Q ₁₂ ", and "Q ₁₃ " is set at the first update timing TP ₁ (1). It is assumed that the fluctuations are "0.1 → 0.2", "0.9 → 1.8", and "0.3 → 0.6", respectively. In this case, the score associated with each search query is doubled due to the update, but when viewed between the scores associated with the search query, there is no change in the relative magnitude relationship. Here, the relative rate of change of other search queries when the rate of change of the score associated with one search query is used as a reference is calculated, and the search query that shows a peculiar change is specified based on the calculated score. You may. More specifically, for example, when the change rate of the score of the search query: "Q ₁₁ " is "1", the change rate of the score associated with the search query: "Q ₁₂ " is "9", and the search query : The rate of change of the score associated with "Q ₁₃ " is "3". If the change rate of the score of the search query: "Q ₁₂ " is "1", the change rate of the score associated with the search query: "Q ₁₁ " is "1/9", and the search query: "Q ₁₃ ". The rate of change in the score associated with "" is "1/3". Also, when the change rate of the score of the search query: "Q ₁₃ " is "1", the change rate of the score associated with the search query: "Q ₁₁ " is "1/3", and the search query: " The rate of change in the score associated with "Q13" is " ₃ ". In this way, the relative change rate of the search query: the score associated with "Q ₁₂ " and the search query: the score associated with "Q ₁₃ " when the change rate of "Q ₁₁ " is used as a reference. Search query: the relative rate of change of the score associated with "Q ₁₁ " and the search query: the relative rate of change of the score associated with "Q ₁₃ " based on the rate of change of "Q ₁₂ ". Search query: Search query based on the rate of change of "Q ₁₃ ": Search query based on the relative rate of change of the score associated with "Q ₁₁ " and search query: Score associated with "Q ₁₂ " : It can be seen that the rate of change in the score associated with "Q12" is generally large. The information processing apparatus 100 can calculate the relative change rate of the score associated with the search query based on the change rate of each search query included in the model M11, and can identify the search query having a relatively large fluctuation in the score. ..

In this way, the information processing apparatus 100 traces the change in the score value for each search query and detects the change over time in the score of the search query. As a result, the operator of the information processing apparatus 100 generates over time in the context of each seed user associated with the predetermined action through the change in the relationship between the search query included in the model and the predetermined action corresponding to the model. You can discover it without overlooking the gradual changes.

In addition, in FIG. 3, the information processing apparatus 100 changes the search model over time between the model M ₁₁ (n + n) generated by the update timing TP ₁ (n + n) and the first model M ₁₁ (0). Although an example of detection is given, the present invention is not particularly limited to this example. For example, the information processing apparatus 100 may detect a change in the score of the search query over time with the model M ₁₁ arbitrarily selected by the operator. This makes it possible to detect changes in the relationship between the search query contained in the model and the predetermined behavior corresponding to the model at various time intervals.

[2. Information processing device configuration example]
Next, an example of the configuration of the information processing apparatus 100 according to the embodiment will be described with reference to FIG. FIG. 4 is a diagram showing a configuration example of the information processing apparatus according to the embodiment. The information processing device 100 shown in FIG. 4 is, for example, a desktop PC (Personal Computer), a notebook PC, a tablet terminal, a smartphone, a mobile phone, a PDA (Personal Digital Assistant), a wearable device, or the like. ..

The information processing device 100 can cooperate with each server device that provides various services through a web page containing various contents, and can acquire various data of service users from each server device. The information processing device 100 can acquire service user data related to various services such as a news site, a technical explanation site, a search engine site, a shopping site, an EC site, an auction site, and a finance site (stock price site) from each server device. .. The service user data that can be acquired by the information processing apparatus 100 includes various search history data. The information processing device 100 may be a device that provides services.

As shown in FIG. 4, the information processing apparatus 100 includes a communication unit 110, a display unit 120, a storage unit 130, and a control unit 140. Note that FIG. 4 shows a configuration example of the information processing apparatus 100, and is not limited to the form shown in FIG. 4, and may be a form including other functional units other than those shown in FIG.

(Communication unit 110)
The communication unit 110 is, for example, connected to a predetermined network by wire or wirelessly, and transmits / receives information to / from another device via the predetermined network. The communication unit 110 is realized by, for example, a NIC (Network Interface Card), an antenna, or the like. The predetermined network is a LAN (Local Area Network), WAN (Wide Area Network), a telephone network (mobile telephone network, fixed telephone network, etc.), a regional IP (Internet Protocol) network, or a communication network such as the Internet. be. The predetermined network may include a wired network or may include a wireless network.

(Display unit 120)
The display unit 120 displays various information. The display unit 120 is a display device. The display unit 120 is realized by a display device such as a liquid crystal display (LCD) or an organic EL display (OELD: Organic Electro-Luminescent Display), for example. The display unit 120 may be realized by a touch panel type display device provided with a touch panel.

(Memory unit 130)
The storage unit 130 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. As shown in FIG. 4, the storage unit 130 includes a search history storage unit 131, a seed user information storage unit 132, and a model storage unit 133.

(Search history storage unit 131)
The search history storage unit 131 stores the search history. FIG. 5 is a diagram showing an example of a search history according to an embodiment. As shown in FIG. 5, the search history stored in the search history storage unit 131 includes a "user ID", a "user name", a "date", a "time", a "search query", and a "search". It has items such as "results".

In the item of "user ID", identification information individually assigned to each user in order to identify the user is stored. In the "user" item, the user name identified by the user ID is stored.

In the "date" item, the date when the user executed the search by the search query is stored. Further, in the "time" item, the time when the user executes the search by the search query is stored. Further, in the item of "search query", the search query used by the user for the search is stored. Further, in the item of "search result", information such as the address of the search result (site) accessed by the user among the search results by the search query is stored.

In the example shown in FIG. 5, the user with the user ID: "UID011" and the user name: "NAME011" has the date: "DATE m * d", the time: "TIME 〇: 〇", and the search query: " _Q11 ". Search result: It is shown that the site of "SITE XYZ" has been accessed.

(Seed user information storage unit 132)
The seed user information storage unit 132 stores information about a plurality of seed users who are users associated with a predetermined action. FIG. 6 is a diagram showing an example of seed user information according to the embodiment.

As shown in FIG. 6, the seed user information stored in the seed user information storage unit 132 includes items of "action ID", "user group ID", "corresponding action ID", and "corresponding user ID". Have.

The item of "user group ID" stores identification information individually assigned to each seed user group in order to identify a seed user group composed of a plurality of seed users who are users associated with a predetermined action. Will be done. For example, the user group ID: "GID 11" shown in FIG. 6 corresponds to the identification information of the seed user group GP ₁₁ shown in FIG. 1 and the like, and the user group ID: "GID 22" shown in FIG. 6 is shown in FIG. 1 and the like. It corresponds to the identification information of the seed user group GP ₂₂ , and the user group ID: “GID 33” shown in FIG. 6 corresponds to the identification information of the seed user group GP ₃₃ shown in FIG. 1 and the like.

In the item of "corresponding action ID", identification information individually assigned to each predetermined action in order to identify the predetermined action is stored. For example, the action ID: “AID_aa” shown in FIG. 6 corresponds to the above-mentioned action AA identification information, and the action ID: “AID_ab” shown in FIG. 6 corresponds to the above-mentioned action AB identification information, and FIG. Action ID: "AID_ac" corresponds to the above-mentioned action AC identification information.

The predetermined action corresponds to an action such as a search using a certain search query, a visit to a certain facility or a store, or a purchase of a certain product, which is arbitrarily set as a target of an advertising strategy or the like. As long as the information can be acquired by the information processing apparatus 100 (or the operator), the predetermined action can be appropriately selected and set.

The seed user can also be rephrased as a user who is highly sensitive to a predetermined behavior (or a user who is sensitive to information about a predetermined behavior). The seed user may be arbitrarily extracted by the operator of the information processing apparatus 100, or may be automatically extracted by the information processing apparatus 100. For example, a user whose frequency of executing a certain action AA exceeds a predetermined threshold value can be regarded as a seed user having high sensitivity to the action AA. For example, when a visit to a certain restaurant is set as a predetermined action, a user who has a high sensitivity to the store, such as visiting the store at least once every two days, can be extracted as a seed user. The frequency that determines the sensitivity to an action may be relatively defined by taking a bird's-eye view of the frequency of the action, or may be arbitrary such as an empirical rule or a social convention for the action. It may be defined by conceptual criteria.

In the item of "corresponding user ID", identification information individually assigned to each user in order to identify a user constituting the corresponding seed user group is stored. This identification information is the same information as the identification information stored in the item of "user ID" of the search history storage unit 131 shown in FIG. As a result, the identification information of the seed user is associated with the search history corresponding to the seed user.

(Model storage unit 133)
The model storage unit 133 stores model information about the model associated with the score indicating the relationship between the seed user and the predetermined action for each search query searched by the seed user. FIG. 7 is a diagram showing an example of model information according to the embodiment. Note that FIG. 7 shows an outline of the model information stored in the model storage unit 133, and may not be configured as shown in FIG. 7.

As shown in FIG. 7, the model information stored in the model storage unit 133 includes items of "model ID", "corresponding action ID", "search query", "score", and "update timing". Have.

In the item of "model ID", identification information individually assigned to each model for identifying a model is stored. The model ID: “MID_1 * 0” shown in FIG. 7 corresponds to the identification information of the model M ₁₁ shown in FIG. 1 and the like, and the model ID: “MID_1 * 0” shown in FIG. 7 corresponds to the model M shown in FIG. 1 and the like. ₂₂ corresponds to the model ID: “MID_3 * 0” shown in FIG. 7, which corresponds to the identification information of the model M ₃₃ shown in FIG. 1 and the like.

In the item of "corresponding action ID", the same information as the information stored in the item of action ID shown in FIG. 6 is stored. That is, in the item of "corresponding action ID", identification information individually allocated for each predetermined action in order to identify the predetermined action is stored. As a result, the seed user information and the model information are associated with each other.

In the item of "search query", the information of the search query used for the search by each user constituting the corresponding seed user group is stored. As the search query information stored in the item of "search query", for example, the search history stored in the search history storage unit 131 is acquired.

In the "score" item, information on the score of the corresponding search query is stored. As the score information stored in the item of "score", the calculated value of the score itself may be stored, or the value obtained by normalizing the calculated value of the score may be stored.

In the item of "update timing", information indicating the update timing for updating the model associated with the score indicating the relationship between the seed user group and the predetermined action is stored for each search query searched by the seed user. This update timing defines the time interval for updating the score for each search query that constitutes the model by generating the model again. In the example shown in FIG. 7, it is shown that the model of the seed user group GP11 to which the user group ID: “GID11” is assigned is updated at the update timing: “TP ₁ ”.

(Control unit 140)
Returning to FIG. 4, the control unit 140 is a controller that controls the information processing device 100. In the control unit 140, various programs (for example, information processing programs) stored in the storage device inside the information processing device 100 are executed by the CPU (Central Processing Unit), MPU (Micro Processing Unit), or the like using the RAM as a work area. It is realized by being done. Further, the control unit 140 may be realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

As shown in FIG. 4, the control unit 140 includes a generation unit 141 and an output unit 142. The control unit 140 realizes or executes the functions and operations of information processing described below by each of these units. The internal configuration of the control unit 140 is not limited to the configuration shown in FIG. 4, and may be any other configuration as long as it is configured to perform information processing described later. Further, the connection relationship of each unit of the control unit 140 is not limited to the connection relationship shown in FIG. 4, and may be another connection relationship. Further, the control unit 140 may include a reception unit when the search history is acquired from the linked server device. Further, the control unit 140 may include a transmission unit when the model is provided to another device.

(Generation unit 141)
The generation unit 141 determines the relationship between the seed user group and the predetermined action for each search query included in the search history based on the search history of the seed user group composed of a plurality of seed users associated with the predetermined action. A model associated with the indicated scores is generated for each predetermined action.

Specifically, the generation unit 141 selects a plurality of users associated with a predetermined action, and sets a seed user group by each selected user. For example, if the predetermined action is a search using a certain search query, the generation unit 141 selects all users whose frequency of executing the search using a certain search query exceeds a predetermined threshold value. When the number of selected users exceeds the preset capacity of the seed user group, the generation unit 141 is a part of the selected users so that the number of seed users constituting the seed user group becomes the capacity, for example. May be arbitrarily truncated. Further, the generation unit 141 does not reach the capacity when the number of selected users is less than the capacity preset in the seed user group, for example, on condition that the number of selected users is a certain number or more. You may set a dummy user for the minute. In addition, for example, when the number of selected users is less than a certain number, the generation unit 141 may change a predetermined behavior and set a seed user group again.

The generation unit 141 stores the set data of the seed user group in the seed user information storage unit 132. Specifically, the generation unit 141 pays out a user group ID individually assigned to the seed user group and an action ID assigned to a predetermined action. Further, the generation unit 141 correlates the user group ID, the action ID, and the user ID of each seed user constituting the seed user group with each other, and stores the user group information in the seed user information storage unit 132 as seed user information. The user ID of each seed user can be obtained from a search system, an EC site, each linked server device, or the like.

After setting the seed user group, the generation unit 141 acquires a search query associated with each seed user constituting the set seed user group from the search history stored in the search history storage unit 131. Further, the generation unit 141 generates a model in which a score indicating the relationship between the seed user group and a predetermined action is associated with each search query based on the acquired search query associated with each seed user. Specifically, the generation unit 141 counts the number of times the seed user inputs each search query input by the user seed, and associates a higher score with the search query as the number of times counted increases. Alternatively, the generation unit 141 may count the number of seed users who input the search query for each search query input by the user seed, and associate a higher score with the search query as the number of times of counting is larger. .. For example, the generation unit 141 can calculate a score indicating the relationship between the seed user group and a predetermined action for each search query included in the search history of the seed user group by the above equation (1). The score calculation method is not limited to the above-mentioned example, and any method may be used as long as the relationship between the seed user group and the predetermined behavior can be evaluated.

In addition, the generation unit 141 updates the model generated for each predetermined action at a predetermined update timing. Specifically, the generation unit 141 refers to the model information stored in the model storage unit 133, and identifies the model at the update timing. Further, the generation unit 141 identifies a seed user group associated with the model, which is the update timing, from the search history stored in the search history storage unit 131. Further, the generation unit 141 acquires a search query associated with each seed user constituting the specified seed user group from the search history stored in the search history storage unit 131. In addition, the generation unit 141 regenerates a model in which a score indicating the relationship between the seed user group and a predetermined action is associated with each search query based on the acquired search query associated with each seed user. , Update the model.

(Output unit 142)
The output unit 142 outputs information based on the fluctuation of the score for each search query each time the model is updated by the generation unit 141. The output unit 142 outputs, for example, a search query in which the rate of increase in the score satisfies a predetermined condition. As a predetermined condition, a condition that the rate of increase in the score is equal to or higher than a predetermined threshold value is exemplified. Further, the output unit 142 outputs, for example, a search query in which the rate of decrease in the score satisfies a predetermined condition. As a predetermined condition, a condition that the rate of decrease in the score is equal to or less than a predetermined threshold value is exemplified. The output unit 142 outputs information based on the fluctuation of the score for each search query to the display unit 120.

[3. Processing procedure]
(3-1. Update process)
Hereinafter, the flow of the update process by the information processing apparatus 100 according to the embodiment will be described with reference to FIG. FIG. 8 is a flowchart showing the flow of update processing by the information processing apparatus according to the embodiment.

As shown in FIG. 8, the information processing apparatus 100 refers to the model information stored in the model storage unit 133 and identifies the model that has reached the update timing (step S101). The information processing apparatus 100 sets the variable i corresponding to the specified model to "1" (step S102).

Further, the information processing apparatus 100 selects one model to be processed from the specified models, and acquires the search history of the seed user group (each seed user constituting) associated with the selected model (). Step S103).

Further, the information processing apparatus 100 refers to the search history of the seed user group, and generates a model in which a score indicating the relationship between the seed user group and a predetermined action is associated with each search query included in the search history. (Step S104).

Further, the information processing apparatus 100 stores the generated model in the model storage unit 133 (step S105).

Further, the information processing apparatus 100 determines whether or not all the models that have reached the update timing have been updated (step S106).

When the information processing apparatus 100 determines that all the models that have reached the update timing have not been updated (step S106; No), the value of the variable i is incremented (“1” is added to the value of the variable i). (Step S107), the process returns to step S103 described above, and the processing procedure shown in FIG. 8 is repeated.

On the other hand, when the information processing apparatus 100 determines that all the models that have reached the update timing have been updated (step S107; Yes), the information processing apparatus 100 ends the processing procedure shown in FIG.

(3-2. Output processing)
Hereinafter, the flow of output processing by the information processing apparatus 100 according to the embodiment will be described with reference to FIG. 9. FIG. 9 is a flowchart showing the flow of output processing by the information processing apparatus according to the embodiment.

As shown in FIG. 9, the output unit 142 identifies the update model updated by the above-mentioned update process (see FIG. 8) (step S201).

Further, the output unit 142 sets "1" in the variable j corresponding to the update model (step S202).

Further, the output unit 142 selects one model to be processed from the update models, and calculates the fluctuation of the score for each search query constituting the selected update model (step S203).

Further, the output unit 142 extracts a search query in which the calculated fluctuation of the score satisfies a predetermined condition (step S204).

Further, the output unit 142 determines whether or not the processing has been completed for all the update models (step S205).

When the output unit 142 determines that the processing has not been completed for all the update models (step S205; No), the output unit 142 increments the value of the variable j (adds "1" to the value of the variable j) (step S206). ), Returning to step S203 described above, the processing procedure shown in FIG. 9 is repeated.

On the other hand, when it is determined that the processing is completed for all the update models (step S205; Yes), the output unit 142 outputs a search query in which the fluctuation of the score satisfies a predetermined condition for each predetermined action (step S207). ), The processing procedure shown in FIG. 9 is completed.

[4. Modification example]
The information processing apparatus 100 according to the embodiment may be implemented in various different forms other than the above-described embodiment. Therefore, a modified example of the information processing apparatus 100 according to the embodiment will be described below.

(4-1. Update timing)
In the above-described embodiment, the information processing apparatus 100 may have different update timings for updating the model for each seed user associated with a predetermined action. FIG. 10 is a diagram showing an example of seed user information according to a modified example.

As shown in FIG. 10, the information processing apparatus 100 updates the model M ₁₁ of the seed user group associated with the action AA at a predetermined timing TP ₁ . Further, the information processing apparatus 100 updates the model M ₂₂ of the seed user group associated with the action AB at a predetermined timing TP ₂ . Further, the information processing apparatus 100 updates the model M ₃₃ of the seed user group associated with the action AC at a predetermined timing TP ₃ . In this way, by making the update timing for updating the model different for each seed user associated with a predetermined behavior, it is possible to output the fluctuation of the score of the search query at the timing according to the behavior.

(4-2. Model generation)
In the above-described embodiment, the generation unit 141 may generate a model by a learning process. For example, the generation unit 141 uses the search history of the seed user who has performed a predetermined action as teacher data, and derives a score indicating the relationship between the search query used by the seed user for the search and the predetermined action by learning processing. .. Specifically, a combination of scores that satisfies the following equation (3) is obtained. In the following equation (3), "y" indicates an objective variable corresponding to the prediction result of whether or not a predetermined action is performed, and "x ₁ " to "x _n " indicate explanatory variables corresponding to the features. , "W ₁ " to "w _n " indicate the respective scores of "x ₁ " to "x _n ". Further, in the following formula (3), "n" indicates a feature number.

The generation unit 141 assigns, for example, a numerical value indicating whether or not the search history of the target user includes a target search query to "x ₁ " to "x _n " in the above equation (3). The generation unit 141 assigns "1" when it is included, and "0" when it is not included. For example, "x ₁ " is assigned information as to whether or not the search history contains the search query: "Q ₁₁ ", and "x ₂ " includes the search query: "Q ₁₂ " in the search history. It is like assigning some information. The number of times the search query is entered may be assigned to "x ₁ " to "x _n " in the above equation (3), or the number of users who have entered the search query may be assigned.

Assuming that the search history shown in FIG. 5 is teacher data, the generation unit 141 assigns a value corresponding to the above equation (3) to the user with the user ID: "UID011", and "1 = w ₁ ". × 1 + w ₂ × 1 + w ₃ × 0 + ... ”. Further, when the value corresponding to the above equation (3) is assigned to the user with the user ID: "UID012", it becomes "1 = w ₁ × 1 + w ₂ × 1 + w ₃ × 1 + ...". In this way, a combination of scores of "w ₁ " to "w _n " that satisfies all the mathematical formulas obtained by allocating the data related to each user to the above-mentioned formula (3) is derived by the learning process.

The algorithms used by the generator 141 for learning processing include classification trees, regression trees, discriminant analysis, k-nearest neighbors, simple bays, support vector machines, and machine learning such as DNN (Deep Neural Network). The algorithm used can be used.

[5. Hardware configuration]
The information processing apparatus 100 according to the embodiment and the modification is realized by a computer 1000 having a configuration as shown in FIG. 11, for example. FIG. 11 is a hardware configuration diagram showing an example of a computer that realizes the functions of the information processing apparatus according to the embodiment and the modified example.

The computer 1000 has a CPU 1100, a RAM 1200, a ROM 1300, an HDD 1400, a communication interface (I / F) 1500, an input / output interface (I / F) 1600, and a media interface (I / F) 1700.

The CPU 1100 operates based on a program stored in the ROM 1300 or the HDD 1400, and controls each part. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started, a program depending on the hardware of the computer 1000, and the like.

The HDD 1400 stores a program executed by the CPU 1100, data used by such a program, and the like. The communication interface 1500 receives data from another device via the network (communication network) 50 and sends it to the CPU 1100, and transmits the data generated by the CPU 1100 to the other device via the network (communication network) 50.

The CPU 1100 controls an output device such as a display or a printer, and an input device such as a keyboard or a mouse via the input / output interface 1600. The CPU 1100 acquires data from the input device via the input / output interface 1600. Further, the CPU 1100 outputs the generated data to the output device via the input / output interface 1600.

The media interface 1700 reads a program or data stored in the recording medium 1800 and provides the program or data to the CPU 1100 via the RAM 1200. The CPU 1100 loads the program from the recording medium 1800 onto the RAM 1200 via the media interface 1700, and executes the loaded program. The recording medium 1800 is, 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 MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory. And so on.

For example, when the computer 1000 functions as the information processing apparatus 100 according to the embodiment, the CPU 1100 of the computer 1000 realizes the function of the control unit 140 by executing the program loaded on the RAM 1200. Further, the data in the storage unit 130 is stored in the HDD 1400. The CPU 1100 of the computer 1000 reads and executes these programs from the recording medium 1800, but as another example, these programs may be acquired from another device via the network (communication network) 50.

[6. others]
Of the processes described in the above-described embodiments and modifications, all or part of the processes described as being automatically performed can be manually performed, or are described as being manually performed. It is also possible to automatically perform all or part of the processing by a known method. In addition, information including processing procedures, specific names, various data and parameters shown in the above documents and drawings can be arbitrarily changed unless otherwise specified.

In the above-described embodiments and modifications, in order to realize the information processing method by the information processing device 100 (see FIGS. 8 and 9), each part (generation unit 141 and output unit 141) of the control unit 140 included in the information processing device 100 is provided. The processing function corresponding to 142) may be realized as an add-on to the information processing program pre-installed in the information processing apparatus 100, or may be flexibly described as a dedicated information processing program using a lightweight programming language or the like. It may be realized by doing.

Further, each component of each of the illustrated devices is a functional concept, and does not necessarily have to be physically configured as shown in the figure. That is, the specific form of distribution / integration of each device is not limited to the one shown in the figure, and all or part of them may be functionally or physically distributed / physically in any unit according to various loads and usage conditions. Can be integrated and configured.

In addition, the above-described embodiments can be appropriately combined as long as the processing contents do not contradict each other.

[7. effect]
The information processing apparatus 100 according to the embodiment includes a generation unit 141 and an output unit 142. The generation unit 141 has a search history of a seed user group (an example of a first user group) composed of a plurality of seed users (an example of a first user) associated with a predetermined action, and a plurality of comparison users other than the seed user. Based on the search history of the second user group (an example of the comparison user group) composed of (an example of the second user), a score indicating the relevance to a predetermined action for each search query included in the search history is given. A associated model is generated for each predetermined action. Further, the generation unit 141 updates the generated model at a predetermined timing. Further, the output unit 142 outputs information based on the fluctuation of the score for each search query each time the model is updated by the generation unit 141.

As described above, the information processing apparatus 100 according to the embodiment generates a model associated with a score indicating a relationship with a predetermined action for each search query, and updates the model at a predetermined timing. Output. Therefore, the information processing apparatus 100 can provide more useful information to a user such as an operator of the information processing apparatus 100. That is, the operator can trace the change in the context of the seed user SU associated with a predetermined action through the change in the score value for each search query, and can detect a peculiar search query that can be a sprout.

Further, the output unit 142 outputs a search query in which the rate of increase in the score satisfies a predetermined condition. Therefore, the information processing apparatus 100 can provide the user with the information of the search query that attracts a lot of attention from the seed user.

Further, the output unit 142 outputs a search query in which the rate of decrease in the score satisfies a predetermined condition. Therefore, the information processing apparatus 100 can provide the user with the information of the search query whose attention of the seed user has decreased.

Further, in the information processing apparatus 100, the seed user is a user who is highly sensitive to a predetermined action, and the comparison user is a user randomly selected from other than the seed user. Therefore, the information processing apparatus 100 can clarify the difference between the search query attracting attention from the seed user and other search queries.

Further, in the information processing apparatus 100, the generation unit 141 makes the timing of updating the model different for each predetermined action. Therefore, the information processing apparatus 100 can output the fluctuation of the score of the search query at the timing according to the action.

The embodiments of the present application have been described in detail with reference to some drawings, but these are examples, and various modifications are made based on the knowledge of those skilled in the art, including the embodiments described in the disclosure column of the invention. It is possible to carry out the present invention in other modified forms.

Further, the above-mentioned "section, module, unit" can be read as "means" or "circuit". For example, the specific unit can be read as a specific means or a specific circuit.

100 Information processing device 110 Communication unit 120 Display unit 130 Storage unit 131 Search history storage unit 132 Seed user information storage unit 133 Model storage unit 140 Control unit 141 Generation unit 142 Output unit

Claims (7)

Based on the search history of a user group composed of a plurality of users associated with a predetermined action, a score indicating the relationship between the user group and the predetermined action is associated with each search query included in the search history. A generation unit that generates a model for each predetermined action and updates it at a predetermined timing.
An information processing device including an output unit that outputs a search query in which the fluctuation of the score for each search query satisfies a predetermined condition each time the model is updated by the generation unit. The information processing device according to claim 1, wherein the output unit outputs the search query in which the rate of increase in the score satisfies a predetermined condition. The information processing apparatus according to claim 1 or 2, wherein the output unit outputs the search query in which the rate of decrease in the score satisfies a predetermined condition. The user
The information processing apparatus according to any one of claims 1 to 3, wherein the user is a user whose frequency of executing the predetermined action exceeds a predetermined threshold value. The generator is
The information processing apparatus according to any one of claims 1 to 4, wherein the timing for updating the model is different for each predetermined action. It is an information processing method executed by a computer.
Based on the search history of a user group composed of a plurality of users associated with a predetermined action, a score indicating the relationship between the user group and the predetermined action is associated with each search query included in the search history. A generation process that generates a model for each predetermined action, and
Each time the model is generated by the generation step, the output step of outputting a search query in which the fluctuation of the score for each search query satisfies a predetermined condition is included.
The generation step is an information processing method for updating the model at a predetermined timing. On the computer
Based on the search history of a user group composed of a plurality of users associated with a predetermined action, a score indicating the relationship between the user group and the predetermined action is associated with each search query included in the search history. A generation procedure for generating a model for each predetermined action, and
Every time the model is generated by the generation procedure, an output procedure for outputting a search query in which the fluctuation of the score for each search query satisfies a predetermined condition is executed.
The generation procedure is an information processing program that updates the model at a predetermined timing.

Images