JP6359045B2 - Information processing apparatus, information processing method, and information processing program - Google Patents

Description

  The present invention relates to an information processing apparatus, an information processing method, and an information processing program.

  For example, there is a technique for classifying a product or service into a plurality of categories in order to predict or learn the occurrence of an event related to the product or service, such as purchase of the product.

  In such a case, a technique for reconstructing the category by integrating the constructed category based on the feature amount of the customer who uses the product or service is provided. In such a technique, categories having similar distributions of feature values that are discrete values are integrated.

JP 2006-318249 A

  However, in the above prior art, the feature values of continuous values are discretized by dividing them into predetermined units. For this reason, for example, due to the influence of rounding error due to discretization, the accuracy of the similarity determination of the distribution of feature amounts may be reduced.

  The present application has been made in view of the above, and an object thereof is to provide an information processing apparatus, an information processing method, and an information processing program capable of improving the integration accuracy of categories.

The information processing apparatus according to the present application is a feature amount of an object included in a plurality of lower categories located below a predetermined category among a plurality of categories having a hierarchical structure, and the feature amount represented by a continuous value based, for each of the plurality of the lower category, the calculation unit for calculating a statistical test value, based on the calculation result of the calculating unit, and integration unit that integrates a plurality of the lower category, Ru comprising a. The integration unit does not integrate the predetermined category with the lower category that is one level lower than the predetermined category when a category exists in the lower level of the predetermined category.

  According to one aspect of the embodiment, the accuracy of category integration can be improved.

FIG. 1 is a diagram illustrating an example of integration processing according to the embodiment. FIG. 2 is a block diagram illustrating a configuration of the information processing apparatus according to the embodiment. FIG. 3A is a diagram illustrating an example of a category information storage unit according to the embodiment. FIG. 3B is a schematic diagram illustrating a hierarchical relationship among a plurality of categories stored in the category information storage unit. FIG. 4 is a diagram illustrating an example of the attribute information storage unit according to the embodiment. FIG. 5A is a graph illustrating an example of a distribution of feature amounts of objects included in a lower category according to the embodiment. FIG. 5B is a graph illustrating an example of a distribution of feature amounts of objects included in a lower category according to the embodiment. FIG. 6 is a diagram illustrating an example of a plurality of categories generated by the integrated processing unit according to the embodiment. FIG. 7A is a graph illustrating an example of a distribution of feature amounts of objects included in a lower category according to the embodiment. FIG. 7B is a graph illustrating an example of a distribution of feature amounts of objects included in a lower category according to the embodiment. FIG. 8A is a graph illustrating an example of a distribution of feature amounts of objects included in a lower category according to the embodiment. FIG. 8B is a graph illustrating an example of a distribution of feature amounts of objects included in a lower category according to the embodiment. FIG. 8C is a graph illustrating an example of a distribution of feature amounts of objects included in a lower category according to the embodiment. FIG. 9 is a diagram illustrating an example of a plurality of categories generated by the integrated processing unit according to the embodiment. FIG. 10 is a diagram illustrating the updated category information storage unit according to the embodiment. FIG. 11 is a diagram illustrating the updated attribute information storage unit according to the embodiment. FIG. 12 is a flowchart illustrating an integration processing procedure by the information processing apparatus according to the embodiment. FIG. 13 is a hardware configuration diagram illustrating an example of a computer that implements the functions of the information processing apparatus.

  Hereinafter, a mode for carrying out an information processing apparatus, an information processing method, and an information processing program according to the present application (hereinafter referred to as “embodiment”) will be described in detail with reference to the drawings. Note that the information processing apparatus, the information processing method, and the information processing program according to the present application are not limited by this embodiment. In the following embodiments, the same portions are denoted by the same reference numerals, and redundant description is omitted.

[1. Integrated processing)
First, an example of the integration process according to the embodiment will be described with reference to FIG. FIG. 1 is a diagram illustrating an example of integration processing according to the embodiment. The integration process according to the embodiment is a process of integrating a plurality of categories having a hierarchical structure, and is performed by an information processing apparatus (not shown).

  The information processing apparatus is an apparatus that integrates categories having a tree structure, for example. The information processing apparatus integrates the categories based on the feature amounts of the objects included in the categories. This point will be described in detail with reference to FIG.

  A category tree T11 shown in FIG. 1 is a classification category tree for classifying objects. Specifically, for example, the object is a product that has been exhibited in the Internet auction, and the classification category tree is a category tree prepared in advance for classifying the product. In the example of FIG. 1, two categories of “desktop type” and “notebook type” are lower categories located below “personal computer”. In addition, the three categories of “product A”, “product B”, and “product C” are lower categories located under the “desktop type”.

  “Product A”, “Product B”, and “Product C” each include, for example, a plurality of products that have been exhibited in an Internet auction (hereinafter referred to as “exhibited products”) as objects. For example, a product ID is allocated to each exhibition product, and the product ID is associated with each category. Hereinafter, the exhibition ID is referred to as a code of each exhibition product. In the example of FIG. 1, the successful bid price of the exhibited product is associated with each product ID as the feature quantity of the exhibited product. The successful bid price of the exhibited product is a continuous variable, and the item value is a continuous value.

  First, the information processing apparatus calculates a statistical test value based on the winning bid price of the exhibited product included in the lower category of the predetermined category (step S1). Here, it is assumed that the predetermined category is “desktop type”. In addition, it is assumed that the information processing apparatus performs t-test. The information processing device calculates a statistical test value that is smaller as the difference between the average values is larger, based on the difference between the average values of the distributions of the feature amounts of the two categories.

As illustrated in FIG. 1, the information processing apparatus calculates a statistical test value for each combination of a plurality of lower categories. The information processing apparatus, for example, combines “product A” and “product B” with “0.60”, “product A”, and “product C” with “1.23e ⁻³⁶ ”, “product B”, “ Assume that a statistical test value of “2.55e ⁻²⁸ ” is calculated for the combination of “Product C”.

  Next, the information processing apparatus integrates a plurality of lower categories based on the calculation result (step S2). Specifically, the information processing apparatus compares the threshold value Th with the calculation result, and integrates a combination of lower categories having a statistical test value larger than the threshold value Th. In FIG. 1, it is assumed that the statistical test value of the combination of “product A” and “product B” is larger than the threshold value Th, and other combinations are equal to or less than the threshold value Th.

  In this case, the information processing apparatus integrates “product A” and “product B” to generate a category tree T12. Here, the information processing apparatus generates a new category including an object included in the category to be integrated below the predetermined category, and deletes the category to be integrated, thereby integrating the categories. In the example of FIG. 1, the information processing apparatus generates a new category “product A and product B” below “desktop type”, and deletes “product A” and “product B”.

  As described above, the information processing apparatus according to the embodiment is a feature value of an object included in a plurality of lower categories located below a predetermined category among a plurality of categories having a hierarchical structure, and is represented by a continuous value. A statistical test value is calculated for each combination of a plurality of subcategories based on the feature quantity. Further, the information processing apparatus integrates a plurality of lower categories based on the calculation result.

  As a result, the information processing apparatus can integrate categories based on feature values that are continuous values. For this reason, the integration accuracy of categories can be improved.

[2. Configuration of information processing apparatus]
The configuration of the information processing apparatus 1 according to the embodiment will be described with reference to FIG. FIG. 2 is a block diagram illustrating a configuration of the information processing apparatus 1 according to the embodiment. As illustrated in FIG. 2, the information processing apparatus 1 includes a storage unit 10 and a control unit (controller) 20.

  The storage unit 10 is realized by, for example, a semiconductor memory device such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 10 includes a category information storage unit 110 and an attribute information storage unit 120.

  The category information storage unit 110 stores category information regarding a plurality of categories having a hierarchical structure. Specifically, the category information storage unit 110 stores hierarchical information that is a hierarchical relationship between categories. Here, an example of the category information storage unit 110 will be described with reference to FIGS. 3A and 3B. FIG. 3A is a diagram illustrating an example of the category information storage unit 110 according to the embodiment. FIG. 3B is a schematic diagram illustrating a hierarchical relationship among a plurality of categories stored in the category information storage unit 110. Here, the plurality of categories will be described as classification categories of the exhibited merchandise as objects, that is, merchandise categories, but are not limited to merchandise categories. Further, although a plurality of categories are described as having a tree structure, any category having a hierarchical structure may be used, and the present invention is not limited to the tree structure.

  As illustrated in FIG. 3A, the category information storage unit 110 includes “parent node” and “child node” as items. The parent node indicates the upper category, and the child node indicates the lower category. In other words, the category of the child node is a lower category located under the category of the parent node. For example, when the parent node is “computer” and the child node is “tablet”, “tablet” is positioned below “computer”.

  FIG. 3B schematically shows a category tree T21 including a plurality of categories stored in the category information storage unit 110. In FIG. 3B, “computer” is positioned as the highest category. “Tablet” and “PC” are located below “Computer”.

  The attribute information storage unit 120 stores information regarding objects included in a plurality of categories. The attribute information storage unit 120 stores attribute information indicating object attributes such as an object name and a category to which the object belongs. Here, an example of the attribute information storage unit 120 will be described with reference to FIG. FIG. 4 is a diagram illustrating an example of the attribute information storage unit 120 according to the embodiment. Here, the description will be made assuming that the object is an exhibition commodity, but the object is not limited to the exhibition commodity as long as it is an object included in the above-described category.

  As shown in FIG. 4, the attribute information storage unit 120 includes items such as “product ID”, “product category”, “successful bid price”, “product image”, “number of bids”, and “product state”. “Product ID” is information indicating a unique number for identifying each product, for example. The “product category” corresponds to the category stored in the category information storage unit 110 and is information indicating in which category the exhibited product is included. The “successful bid price” and “number of bids” are continuous variables indicating the price at which the exhibited product is awarded and the number of bids, respectively. “Product image” and “product state” are discrete variables indicating the presence and state of an image.

  Returning to FIG. The control unit 20 includes, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like that stores various programs (corresponding to an example of an information processing program) stored in the storage unit 10 inside the information processing apparatus 1 in RAM. Is implemented as a work area. The control unit 20 is realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

  The control unit 20 includes a processing unit 200, a feature designation unit 210, an acquisition unit 220, a calculation unit 230, and an integration unit 240, and realizes or executes information processing functions and operations described below. Note that the internal configuration of the control unit 20 is not limited to the configuration illustrated in FIG. 2, and may be other configurations as long as the configuration performs information processing to be described later. Further, the connection relationship between the processing units included in the control unit 20 is not limited to the connection relationship illustrated in FIG. 2, and may be another connection relationship.

  The processing unit 200 performs processing using the attribute information stored in the storage unit 10. For example, when a user considers whether or not to sell a predetermined product, there is a case where it is desired to predict how much a product scheduled to be sold (hereinafter referred to as a scheduled product to be sold) will be sold. In this case, the processing unit 200 performs a prediction model calculation using the attribute information as a feature amount.

  However, for example, when the number of products in the same category as the planned product for sale is small, the accuracy of the prediction calculation by the generated model may be reduced. Therefore, in the information processing apparatus 1 according to the embodiment, for example, in order to improve calculation accuracy using attribute information, a plurality of categories are integrated. At this time, by integrating the categories based on the feature amount corresponding to the prediction calculation processing by the processing unit 200, the categories can be integrated without reducing the prediction calculation accuracy by the processing unit 200.

  The feature designation unit 210 designates a feature amount used for calculating the statistical test value calculated by the calculation unit 230. The feature amount designated by the feature designation unit 210 varies depending on the processing performed by the processing unit 200. For example, when the processing unit 200 performs a bid price prediction process, the feature designation unit 210 designates “successful bid price” as the feature amount. The feature quantity designated by the feature designation unit 210 may be a feature quantity that is a continuous value among the feature quantities included in the attribute information stored in the attribute information storage unit 120, and is not limited to the “successful bid price”. For example, in the example of FIG. 4, the feature designation unit 210 can also designate “the number of bids”.

  Here, a case has been described in which the feature specifying unit 210 specifies a feature amount according to the calculation by the processing unit 200, but the present invention is not limited to this. For example, a feature amount designated in advance may be determined. In this case, for example, information regarding the feature quantity to be designated is stored in the storage unit 10, and the feature designation unit 210 designates the feature quantity with reference to the storage unit 10. Alternatively, the user may specify the feature amount. In this case, the feature designation unit 210 functions as a reception unit, for example. The feature designation unit 210 designates the feature amount received from the user. The feature designation unit 210 outputs the designated feature amount to the acquisition unit 220 and the calculation unit 230.

  The acquisition unit 220 acquires hierarchical information of a plurality of categories. Further, the acquisition unit 220 acquires the feature amount of the object included in the category. Specifically, the acquisition unit 220 acquires category information from the category information storage unit 110. The acquisition unit 220 acquires the feature amount designated by the feature designation unit 210 from the attribute information storage unit 120. The acquisition unit 220 outputs the acquired category information and feature amount to the calculation unit 230.

  The calculation unit 230 calculates a statistical test value based on the category information and the feature amount acquired by the acquisition unit 220. Specifically, the calculation unit 230 calculates a statistical test value for each combination of a plurality of lower categories located below the predetermined category. The calculation unit 230 outputs the calculated statistical test value to the integration unit 240.

  The integration unit 240 integrates the categories based on the statistical test value calculated by the calculation unit 230. The integration unit 240 includes an integration determination unit 241 and an integration processing unit 242.

  The integration determining unit 241 determines a category to be integrated. For example, when one lower category is positioned below the predetermined category, the integration determining unit 241 determines to integrate the lower category and the predetermined category. Further, the integration determining unit 241 determines a category to be integrated according to the statistical test value calculated by the calculating unit 230. For example, the integration determining unit 241 determines a category for integrating a combination of lower categories whose statistical test value is equal to or greater than a predetermined threshold Th. The integration determining unit 241 outputs the determined combination of categories to the integration processing unit 242.

  The integration processing unit 242 integrates the combination of categories determined by the integration determination unit 241 and generates a new category. The integration processing unit 242 updates the category information stored in the category information storage unit 110 based on the generated category. Further, based on the generated category, the integration processing unit 242 updates the attribute information stored in the attribute information storage unit 120. For example, it is assumed that the information processing apparatus 1 performs category integration processing until there is no category that can be integrated, based on the category information and attribute information updated by the integration processing unit 242.

  Next, details of the integration process performed by the information processing apparatus 1 will be described with reference to FIGS. 5A to 11. Here, it is assumed that the integration process is performed on a plurality of categories shown in FIGS. 3A and 3B. First, the calculation unit 230 of the information processing apparatus 1 sets the upper category of the lowest category described as the classification category in the attribute information as a predetermined category. In addition, the calculation unit 230 sets one of the predetermined categories to which the lowest category belongs, as the designated category. In the example of FIGS. 3A and 3B, the predetermined categories are “brand A”, “company B”, “brand C” to “brand E”. The calculation unit 230 sets “brand E” as the designated category from the predetermined categories.

  The calculation unit 230 calculates a statistical test value for each combination of lower categories of the designated category. In the example of FIGS. 3A and 3B, the calculation unit 230 calculates a statistical test value based on the “successful bid price” for the combination of “product E1” and “product E2”. Here, it is assumed that the calculation unit 230 performs a t test and calculates “1” when the average values of the two categories are equal, and calculates a statistical test value that approaches 0 as the difference between the average values increases.

  The integration determining unit 241 of the information processing apparatus 1 determines a combination of a predetermined category having one lower category and a category integrating the lower category. In the example of FIGS. 3A and 3B, the integration determining unit 241 performs a combination of “Brand A” and “Product A”, a combination of “B Company” and “Product B”, and a combination of “Brand C” and “Product C”. , And the combination of “brand D” and “product D” is determined as a combination of categories.

  Further, the integration determination unit 241 determines a combination of categories to be integrated based on the calculation result of the calculation unit 230. For example, it is assumed that the “successful bid price” of the exhibition product included in “product E1” has the distribution shown in FIG. 5A, and the “successful bid price” of the exhibition product included in “product E2” has the distribution shown in FIG. 5B. 5A and 5B are graphs showing an example of the distribution of the feature amounts of the objects included in the lower category, in which the vertical axis indicates the number of products and the horizontal axis indicates the price.

  As shown in FIGS. 5A and 5B, it is assumed that the distribution of “product E1” and the distribution of “product E2” are similar. In this case, the difference between the average values of the distributions of the two categories becomes small, and the calculation unit 230 calculates a t-test value close to “1”. The integration determining unit 241 determines a combination of categories for integrating the combination of “product E1” and “product E2”.

  Note that when “product E1” and “product E2” are integrated, one category is positioned below the designated category “brand E”. In such a case, the integration determining unit 241 determines that the designated category “brand E” is also integrated into “product E1” and “product E2”.

  The integration processing unit 242 of the information processing apparatus 1 integrates the combination of categories determined by the integration determination unit 241. Specifically, the integration processing unit 242 generates a new category and includes an object included in the category determined to be integrated into the new category. At this time, for example, when the category to be integrated includes a higher category, the integration processing unit 242 sets the higher category name as a new category name. In other words, the lower category is included in the upper category, and the lower category is deleted. Further, when the upper category is not included in the category to be integrated, for example, the integration processing unit 242 concatenates the lower category names to form a new category name.

  FIG. 6 shows a plurality of categories updated by the integration processing unit 242 as a tree T22. FIG. 6 is a diagram illustrating an example of a plurality of categories generated by the integration processing unit 242 according to the embodiment. As shown in FIG. 6, the integrated processing unit 242 displays “Product A” for “Brand A”, “Product B” for “Company B”, “Product C” for “Brand C”, and “Brand D”. "Product D" is integrated into each. Further, the integration processing unit 242 integrates “product E1” and “product E2” with “brand E”. Thereby, the integration processing unit 242 updates the category information.

  The information processing apparatus 1 further performs integration processing on a plurality of categories generated by the integration processing unit 242. In this case, for example, the calculation unit 230 calculates a statistical test value for the combination of “brand A” and “company B”. In addition, the calculation unit 230 calculates a statistical test value for all combinations of “brand C”, “brand D”, and “brand E”.

  The integration determination unit 241 determines a combination of categories to be integrated based on the calculation result of the calculation unit 230. For example, it is assumed that the “successful bid price” of the exhibited product included in “brand A” has the distribution shown in FIG. 7A, and the “successful bid price” of the exhibited product included in “B company” has the distribution shown in FIG. 7B. 7A and 7B are graphs showing an example of the distribution of feature amounts of objects included in the lower category, in which the vertical axis indicates the number of products and the horizontal axis indicates the price.

  As shown in FIGS. 7A and 7B, it is assumed that the distribution of “Brand A” and the distribution of “Company B” are similar. In this case, the difference between the average values of the distributions of the two categories becomes small, and the calculation unit 230 calculates a statistical test value close to “1”. In this case, the integration determining unit 241 determines a combination of categories for integrating the combination of “brand A” and “company B”.

  When “Brand A” and “Company B” are integrated, there is one category located under the designated category “Tablet”. In such a case, the integration determining unit 241 determines that the designated category “tablet” is also integrated into “brand A” and “company B”.

  Further, the integration determining unit 241 determines a combination of categories to be integrated from “brand C”, “brand D”, and “brand E”. For example, the “successful bid price” of the exhibition product included in “brand C” has the distribution shown in FIG. 8A, the “successful bid price” of the exhibition product included in “brand D” has the distribution shown in FIG. Suppose that the “successful bid price” of the exhibited products included in the distribution shown in FIG. 8C. 8A to 8C are graphs showing examples of distributions of feature amounts of objects included in the lower category, in which the vertical axis indicates the number of products and the horizontal axis indicates the price. Further, the distribution of “successful bid price” of “brand E” shown in FIG. 8C is a distribution in which the distribution of “successful bid price” of “product E1” and “product E2” shown in FIGS. 5A and 5B is combined. .

  As shown in FIGS. 8A to 8C, the distribution of “Brand C” and the distribution of “Brand D” are similar, and the distribution of “Brand E” is similar to the distribution of “Brand C” and “Brand D”. Shall not. In this case, the difference between the average values of the distributions of “brand C” and “brand D” becomes small, and the calculation unit 230 calculates a statistical test value close to “1”. On the other hand, the difference between the average values of the distributions of “Brand C” and “Brand E” and the distributions of “Brand D” and “Brand E” is large, and the calculation unit 230 calculates a statistical test value close to “0”, respectively. To do. In this case, the integration determining unit 241 determines a combination of categories for integrating a combination of “brand C” and “brand D”. “Brand E” is determined not to be integrated.

  The integration processing unit 242 of the information processing apparatus 1 integrates the combination of categories determined by the integration determination unit 241. FIG. 9 shows the category tree T23 after the integration processing unit 242 integrates the categories. FIG. 9 is a diagram illustrating an example of a plurality of categories generated by the integration processing unit 242 according to the embodiment. As illustrated in FIG. 9, the integration processing unit 242 integrates “brand A” and “company B” into “tablet”. Further, the integration processing unit 242 integrates “brand C” and “brand D” by generating “brand C and D”.

  Here, as shown in FIG. 9, “Brand C and D” and “Brand E” are two levels below “Computer”. Therefore, “tablet” and “personal computer”, which are subcategories of “computer”, cannot be integrated. “Tablet” has no lower category, and “Brand E”, which is a lower category of “PC”, is a category that the integration determining unit 241 determines not to integrate. Thus, the category tree T23 shown in FIG. 9 does not have a category that can be integrated. In this case, the integration processing unit 242 of the information processing device 1 updates the category information and the attribute information, and ends the integration process.

  FIG. 10 is a diagram illustrating the updated category information storage unit 110 according to the embodiment. As illustrated in FIG. 10, the integration processing unit 242 deletes “product A” to “product E2”, “brand A”, and “company B”, and updates the category information. FIG. 11 is a diagram illustrating the updated attribute information storage unit 120 according to the embodiment. As illustrated in FIG. 11, the integration processing unit 242 changes the “product category” of the attribute information to the integrated category based on the updated category information. In FIG. 11, the integration processing unit 242 changes, for example, the “product category” of the product A1 from “product A” to “tablet” (see FIG. 4).

[3. (Integration procedure)
Next, a procedure of integration processing by the information processing apparatus 1 according to the embodiment will be described with reference to FIG. FIG. 12 is a flowchart illustrating an integration processing procedure by the information processing apparatus 1 according to the embodiment. The information processing apparatus 1 according to the embodiment performs the following integration process on a category (predetermined category) one level higher than the lower category. For example, the integration process is repeatedly performed until there is no category that can be integrated.

  As illustrated in FIG. 12, the information processing apparatus 1 designates one of the predetermined categories as a designated category (step S101). The information processing apparatus 1 determines whether a category exists in the lower level of the designated category (step S102). If a category exists in the lower level (step S102; Yes), the information processing apparatus 1 proceeds to step S108 on the assumption that no integration is performed. On the other hand, if there are no two lower categories, that is, if all the lower categories of the designated category are the lowest categories (step S102; No), the information processing apparatus 1 has one lower category number of the designated category. Whether or not (step S103).

  When the number of lower categories is 1, that is, one lower category belongs to the designated category (step S103; Yes), the information processing apparatus 1 integrates the lower category and the designated category (step S104), and proceeds to step S108.

  On the other hand, when the number of lower categories is not 1, that is, when a plurality of lower categories belong to the designated category (step S103; No), the information processing apparatus 1 calculates a statistical test value for all combinations of the plurality of lower categories. (Step S105). The information processing apparatus 1 determines a lower category to be integrated based on the statistical test value calculated in step S105 (step S106), and integrates the determined lower category (step S107).

  The information processing apparatus 1 determines whether or not the integration determination has been performed for all the predetermined categories (step S108). If there is a predetermined category that has not been subjected to the integrated determination of the lower category (step S108; No), the information processing apparatus 1 changes the predetermined category (undetermined category) that has not been subjected to the integrated determination to the designated category (step S109). Return to step S102. On the other hand, when the integration determination is performed for all the predetermined categories (step S108; Yes), the information processing apparatus 1 ends the integration process.

  Thereby, the information processing apparatus 1 can improve the integration accuracy of a category. Therefore, the number of categories can be reduced. For example, when processing using a category such as prediction calculation is performed, the processing speed can be improved without reducing the processing accuracy.

[4. (Modification)
The information processing apparatus 1 according to the above-described embodiment may be implemented in various different forms other than the above-described embodiment. Accordingly, other embodiments will be described below.

  In the above embodiment, the information processing apparatus 1 performs the t test and calculates the statistical test value, but is not limited thereto. The information processing apparatus 1 may perform a statistical test based on feature values that are continuous values. For example, the Brunner-Munzel test or the Mann-Whitney U test may be performed. Further, a Welch t test may be performed as a t test.

  In the above-described embodiment, the information processing apparatus 1 repeatedly performs integration processing until it is determined that integration cannot be performed for all categories, but is not limited thereto. For example, the integration process may be repeated according to the number of objects included in the lower category. That is, the integration unit 240 of the information processing apparatus 1 integrates lower categories in which the number of objects is a predetermined value or less. Specifically, the integration determining unit 241 of the information processing device 1 determines a lower category to be integrated with a lower category in which the number of objects is a predetermined value or less based on the statistical test value calculated by the calculating unit 230.

  For example, in the arithmetic processing using the category information by the processing unit 200, when the number of objects included in the lower category before integration is equal to or less than a predetermined value, the number of objects is small and the arithmetic processing accuracy is reduced. Even in such a case, the number of objects can be increased to a predetermined value or higher by integrating the lower categories having the number of objects equal to or smaller than the predetermined value with high accuracy, and the arithmetic processing accuracy can be improved. Note that the predetermined value is, for example, a number that can be statistically tested by the calculation unit 230 with a predetermined accuracy. Alternatively, the predetermined value may be a number that can be calculated with a predetermined accuracy in the arithmetic processing by the processing unit 200, for example.

  Further, the integration unit 240 of the information processing apparatus 1 may integrate lower categories according to the number of categories. That is, the integration unit 240 of the information processing apparatus 1 integrates lower categories according to the total number of categories.

  For example, in the arithmetic processing using the category information by the processing unit 200, if the total number of categories included in the category information is large, the calculation accuracy is improved, but the calculation amount is also increased. Therefore, the information processing apparatus 1 integrates the lower categories so that, for example, the number of categories according to the calculation amount that can be processed by the processing unit 200 and the required calculation accuracy are obtained. As a result, the categories can be integrated so as to have a desired number of categories according to the arithmetic processing of the processing unit 200.

  Moreover, in the said embodiment, although the process part 200 of the information processing apparatus 1 performed the calculation process of the prediction model of "successful bid price", it is not limited to this. The processing unit 200 only needs to perform processing based on category information and attribute information. For example, a prediction model for predicting whether or not the user bids, a prediction model for predicting whether or not there is a violation in the listing of products, etc. Various calculations may be performed.

  Moreover, in the said embodiment, although the information processing apparatus 1 was set as the structure provided with the process part 200, it is not limited to this. For example, another device different from the information processing device 1 may include the processing unit 200 and perform calculations using category information. In this case, the information processing apparatus 1 integrates a plurality of categories in response to a request from another apparatus.

  Moreover, in the said embodiment, although the information processing apparatus 1 memorize | stores category information and attribute information, it is not limited to this. The category information and attribute information may be stored in a storage device or server device (not shown). In this case, it is assumed that the acquisition unit 220 of the information processing apparatus 1 acquires category information and attribute information from a storage device, a server device, or the like.

  Of the processes described in the above embodiment, all or part of the processes described as being automatically performed can be performed manually, or all of the processes described as being performed manually or A part can be automatically performed by a known method. In addition, the processing procedures, specific names, and information including various data and parameters shown in the document and drawings can be arbitrarily changed unless otherwise specified. For example, the various types of information illustrated in each drawing is not limited to the illustrated information.

  Further, each component of each illustrated apparatus is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured. For example, the feature designation unit 210 and the acquisition unit 220 illustrated in FIG. 2 may be integrated.

  In the above-described embodiment, the attribute information is information related to auctioned products for Internet auction, but is not limited thereto. The attribute information may be information related to a product sold via the Internet, for example. In this case, the category information is a product category for classifying products to be sold. Further, the attribute information may be information including continuous feature values, and may be information about the user, for example. As described above, the attribute information is not limited to information about the product.

[5. Hardware configuration)
Further, the information processing apparatus 1 according to the embodiment described above is realized by a computer 1000 having a configuration as shown in FIG. 13, for example. FIG. 13 is a hardware configuration diagram illustrating an example of a computer 1000 that implements the functions of the information processing apparatus 1. The computer 1000 includes a CPU 1100, RAM 1200, ROM 1300, HDD 1400, communication interface (I / F) 1500, input / output interface (I / F) 1600, and 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 unit. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started up, a program depending on the hardware of the computer 1000, and the like.

  The HDD 1400 stores programs executed by the CPU 1100, data used by the programs, and the like. The communication interface 1500 receives data from other devices via the communication network 500 and sends the data to the CPU 1100, and transmits data generated by the CPU 1100 to other devices via the communication network 500.

  The CPU 1100 controls an output device such as a display and a printer and an input device such as a keyboard and a mouse via the input / output interface 1600. The CPU 1100 acquires data from the input device via the input / output interface 1600. In addition, 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 it to the CPU 1100 via the RAM 1200. The CPU 1100 loads such a 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 an MO (Magneto Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory. It is.

  For example, when the computer 1000 functions as the information processing apparatus 1 according to the embodiment, the CPU 1100 of the computer 1000 implements the function of the control unit 20 by executing a program loaded on the RAM 1200. The HDD 1400 stores data in the storage unit 10. The CPU 1100 of the computer 1000 reads these programs from the recording medium 1800 and executes them, but as another example, these programs may be acquired from other devices via the communication network 500.

[6. effect〕
As described above, the information processing apparatus 1 according to the embodiment includes the calculation unit 230 and the integration unit 240. The calculation unit 230 is a feature amount of an object included in a plurality of lower categories located below a predetermined category among a plurality of categories having a hierarchical structure, and is based on a feature amount represented by a continuous value. A statistical test value is calculated for each combination of subcategories. The integration unit 240 integrates a plurality of lower categories based on the calculation result of the calculation unit 230.

  As a result, the information processing apparatus 1 can integrate categories based on feature values that are continuous values. For this reason, the integration accuracy of categories can be improved. For this reason, the number of categories can be reduced. For example, when processing using a category such as prediction calculation is performed, the processing speed can be improved without reducing the processing accuracy.

  Further, the integration unit 240 according to the embodiment integrates lower categories in which the number of objects included in the lower categories is equal to or less than a predetermined value. As a result, the lower categories whose number of objects is equal to or less than the predetermined value can be integrated with high accuracy, and the number of objects can be increased. For this reason, when processing using categories, such as prediction calculation, is performed, processing accuracy can be improved, for example.

  Further, the integration unit 240 according to the embodiment integrates lower categories according to the number of categories. As a result, the categories can be integrated so that the desired number of categories is obtained. For this reason, it is possible to integrate the categories so that the total number of the plurality of categories becomes the number of categories according to the processing accuracy and the processing speed when performing processing using categories such as prediction calculation.

  The information processing apparatus 1 according to the embodiment further includes an acquisition unit 220 that acquires hierarchical information (category information) of a plurality of categories. Further, the calculation unit 230 calculates a statistical test value for each combination of lower categories based on the hierarchy information acquired by the acquisition unit 220. Thereby, a plurality of categories that are hierarchized in advance can be integrated with high accuracy.

  The object according to the embodiment is a product, and the category is a product category related to the product. Further, the calculation unit 230 calculates a statistical test value using the attribute information of the product as a feature amount. Thereby, the information processing apparatus 1 can integrate each category of the product category with high accuracy.

  As described above, some of the embodiments of the present application have been described in detail with reference to the drawings. However, these are merely examples, and various modifications, including the aspects described in the disclosure section of the invention, based on the knowledge of those skilled in the art, It is possible to implement the present invention in other forms with improvements.

DESCRIPTION OF SYMBOLS 1 Information processing apparatus 110 Category information storage part 120 Attribute information storage part 200 Processing part 220 Acquisition part 230 Calculation part 240 Integration part 241 Integration determination part 242 Integration processing part

Claims (7)

Among a plurality of categories having a hierarchical structure, feature quantities of objects included in a plurality of lower categories located below a predetermined category, and based on the feature quantities represented by continuous values, For each combination, a calculation unit that calculates a statistical test value,
An integration unit that integrates a plurality of the lower categories based on a calculation result of the calculation unit;
Equipped with a,
The information processing apparatus according to claim 1, wherein the integration unit does not integrate the predetermined category with the lower category that is one level lower than the predetermined category when a category exists in the lower level of the predetermined category . The integration unit
Integrating the lower categories in which the number of the objects included in the lower category is a predetermined value or less;
The information processing apparatus according to claim 1. The integration unit
Consolidating the subcategories according to the number of the plurality of categories,
The information processing apparatus according to claim 1, wherein the information processing apparatus is an information processing apparatus. An acquisition unit for acquiring a plurality of hierarchical information of the category;
The calculation unit includes:
Based on the hierarchy information acquired by the acquisition unit, to calculate the statistical test value for each combination of the lower categories,
The information processing apparatus according to any one of claims 1 to 3. The object is a product, and the category is a product category related to the product,
The calculation unit includes:
Calculating the statistical test value using the product attribute information as the feature amount;
The information processing apparatus according to claim 1, wherein the information processing apparatus is an information processing apparatus. Among a plurality of categories having a hierarchical structure, a feature amount of an object included in a plurality of lower categories positioned below a predetermined category, and a combination of the plurality of lower categories based on a feature amount represented by a continuous value A calculation step for calculating a statistical test value for each,
An integration step of integrating a plurality of the lower categories based on a calculation result in the calculation step;
Only including,
In the information processing method , the integrating step does not integrate the predetermined category with the lower category that is one level lower than the predetermined category when a category exists two levels below the predetermined category . On the computer,
Among a plurality of categories having a hierarchical structure, a feature amount of an object included in a plurality of lower categories positioned below a predetermined category, and a combination of the plurality of lower categories based on a feature amount represented by a continuous value For each, a calculation procedure for calculating a statistical test value,
An integration procedure for integrating a plurality of the lower categories based on a calculation result in the calculation procedure;
Was executed,
In the information processing program , the integration procedure does not integrate the predetermined category with the lower category that is one level lower than the predetermined category when a category exists in the lower level of the predetermined category .

Images