Information processing device, classification method, and classification program

Abstract

In order to automatically classify data without using a classifier built through machine learning, an information processing device (1) is provided with: a data acquisition unit (11) which acquires target data that is to be classified into one of a plurality of categories in a hierarchical structure; and a classification unit (12) which classifies the target data into one of the categories on the basis of a matching degree indicating the degree of matching of the target data to a category and higher-level matching degrees indicating the degrees of matching of the target data to categories at a level higher than said category.

Description

[Technical field]

[0001] The present invention relates to an information processing device or the like that classifies data to be classified into categories. [Background technology]

[0002] In recent years, as a large amount of various data has been collected and accumulated, the cost required for classifying the accumulated data to effectively use it has also increased. As a technology for reducing such costs, for example, Patent Document 1 listed below can be mentioned. Patent Document 1 listed below discloses an information processing device that classifies product data related to products or services sold via a network into various categories.

[0003] More specifically, the information processing device described in Patent Document 1 uses product data classified into hierarchical categories as learning data and determines a category using a classifier trained to output a hierarchical category classification result for a product indicated by input product data. This information processing device can automatically classify product data, thereby reducing the human cost required for classifying product data. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2019-164402 Summary of the Invention [Problem to be solved by the invention]

[0005] However, when using a classifier constructed by machine learning as in Patent Document 1, there is a problem that a highly accurate classification result cannot be output unless there is sufficient learning data for each category. One aspect of the present invention aims to provide an information processing device or the like that can automatically classify data without using a classifier constructed by machine learning. [Means for solving the problem]

[0006] An information processing device according to one aspect of the present invention comprises a data acquisition means for acquiring target data to be classified into one of a plurality of categories in a hierarchical structure, and a classification means for classifying the target data into one of the plurality of categories based on a suitability indicating the degree to which the target data fits into the category and a higher-level suitability indicating the degree to which the target data fits into a higher-level category of the category.

[0007] A classification method according to one aspect of the present invention includes at least one processor acquiring target data to be classified into one of a plurality of categories in a hierarchical structure, and classifying the target data into one of the plurality of categories based on a suitability indicating the degree to which the target data fits into the category and a higher-level suitability indicating the degree to which the target data fits into a higher-level category of the category.

[0008] A classification program according to one aspect of the present invention causes a computer to function as a data acquisition means for acquiring target data to be classified into one of a plurality of categories in a hierarchical structure, and a classification means for classifying the target data into one of the plurality of categories based on a suitability indicating the degree to which the target data fits into the category and a higher-level suitability indicating the degree to which the target data fits into a higher-level category of the category. Effect of the Invention

[0009] According to one aspect of the present invention, data can be automatically classified without using a classifier constructed by machine learning. [Brief description of the drawings]

[0010] [Figure 1] 1 is a block diagram showing a configuration of an information processing device according to a first exemplary embodiment of the present invention. [Diagram 2] FIG. 2 is a flow chart showing the flow of a classification method according to the first exemplary embodiment of the present invention. [Diagram 3] FIG. 11 is a block diagram showing a configuration of an information processing device according to a second exemplary embodiment of the present invention. [Figure 4] 11 is a diagram illustrating an example of calculation of an overall compatibility degree by the information processing device. [Diagram 5] FIG. 4 is a diagram illustrating a classification method executed by the information processing device. [Figure 6] FIG. 11 is a block diagram showing a configuration of an information processing device according to an exemplary embodiment 3 of the present invention. [Figure 7] 11 is a diagram illustrating an example of calculation of a degree of compatibility based on a degree of similarity between web pages detected in a web search by the information processing device. [Figure 8] 4 is a flowchart showing a flow of a classification method executed by the information processing device. [Figure 9] FIG. 1 is a diagram illustrating an example of a computer that executes instructions of a program, which is software that realizes each function of an information processing device according to each exemplary embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Example Embodiment 1 A first exemplary embodiment of the present invention will be described in detail with reference to the drawings. This exemplary embodiment is a basic form of the exemplary embodiments described later.

[0012] (Configuration of information processing device 1) The configuration of an information processing device 1 according to this exemplary embodiment will be described with reference to Fig. 1. Fig. 1 is a block diagram showing the configuration of the information processing device 1. As shown in Fig. 1, the information processing device 1 includes a data acquisition unit 11 and a classification unit 12.

[0013] The data acquisition unit 11 acquires target data to be classified into any one of a plurality of categories in a hierarchical structure.

[0014] The classification unit 12 classifies the target data into one of the multiple categories based on a suitability indicating the degree to which the target data matches the category and a higher-level suitability indicating the degree to which the target data matches a higher-level category of the category.

[0015] As described above, the information processing device 1 according to this exemplary embodiment is configured to include a data acquisition unit 11 that acquires target data to be classified into one of a plurality of categories in a hierarchical structure, and a classification unit 12 that classifies the target data into one of a plurality of categories based on a suitability indicating the degree to which the target data fits into the category and a higher-level suitability indicating the degree to which the target data fits into a higher-level category of the category.

[0016] When categories have a hierarchical structure, if target data can be classified into the correct category, the degree of conformity between the target data and the higher-level category is often high. For example, assume that the product name "Tapioca Tea" is the target data. The correct classification of this target data is the higher-level category "tea" and the lower-level category "tapioca milk tea." In this case, "tapioca tea" has a high degree of conformity with "tapioca milk tea" and also with "tea."

[0017] Therefore, according to the above configuration, in which the target data is classified based not only on its suitability for a certain category but also on its suitability for a higher-level category, it becomes possible to classify the target data into an appropriate category even in cases where an appropriate category cannot be identified from its suitability for the category alone.

[0018] For example, in the above example, suppose that a classification called "tapioca sour" exists in the lower category. In this case, it is possible that there is no difference in the suitability between "tapioca tea" and "tapioca sour" and the suitability between "tapioca tea" and "tapioca milk tea", or the suitability of "tapioca sour" is higher. Even in such a case, if the higher category of "tapioca sour" is, for example, "alcohol", it is considered that the suitability of "tapioca tea" with respect to "alcohol" is lower than the suitability of "tapioca tea" with respect to "tea". Therefore, by classifying based on the suitability with respect to the higher category, it is possible to correctly classify "tapioca tea" as "tapioca milk tea".

[0019] In addition, according to the above configuration, it is not necessary to use a classifier constructed by machine learning. In this way, according to the above configuration, it is possible to obtain an effect that it is possible to automatically classify target data without using a classifier constructed by machine learning.

[0020] (Conversion pattern determination program) The above-mentioned functions of the information processing device 1 can also be realized by a program. The classification program according to this exemplary embodiment is configured to cause a computer to function as a data acquisition means for acquiring target data to be classified into one of a plurality of categories in a hierarchical structure, and a classification means for classifying the target data into one of a plurality of categories based on a suitability indicating the degree to which the target data fits the category and a higher suitability indicating the degree to which the target data fits a higher category than the category. Therefore, according to the classification program according to this exemplary embodiment, it is possible to obtain an effect that the target data can be automatically classified without using a classifier constructed by machine learning.

[0021] (Classification method flow) The flow of the classification method according to this exemplary embodiment will be described with reference to Fig. 2. Fig. 2 is a flow diagram showing the flow of the classification method. Note that the execution subject of each step in this classification method may be a processor included in the information processing device 1, or may be a processor included in another device, or each step may be a processor provided in a different device.

[0022] In S11, at least one processor acquires target data to be classified into any one of a plurality of categories in a hierarchical structure.

[0023] In S12, at least one processor classifies the target data into one of the multiple categories based on a suitability indicating the degree to which the target data fits into the category and a higher-level suitability indicating the degree to which the target data fits into a higher-level category of the category.

[0024] As described above, the classification method according to the present exemplary embodiment includes a configuration in which at least one processor acquires target data to be classified into one of a plurality of categories in a hierarchical structure, and classifies the target data into one of the plurality of categories based on a suitability indicating the degree to which the target data fits the category and a higher suitability indicating the degree to which the target data fits a higher category than the category. Therefore, the classification method according to the present exemplary embodiment has an effect of automatically classifying target data without using a classifier constructed by machine learning.

[0025] Exemplary embodiment 2 (Configuration of information processing device 2) The configuration of the information processing device 2 will be described with reference to Fig. 3. Fig. 3 is a block diagram showing the configuration of the information processing device 2. As shown in the figure, the information processing device 2 includes a control unit 20 that controls each unit of the information processing device 2, and a storage unit 21 that stores various data used by the information processing device 2. The information processing device 2 also includes a communication unit 22 that allows the information processing device 2 to communicate with other devices, an input unit 23 that accepts input of various data to the information processing device 2, and an output unit 24 that allows the information processing device 2 to output various data.

[0026] The control unit 20 includes a data acquisition unit 201, a classification data acquisition unit 202, a hierarchical structure specification unit 203, a degree of conformance calculation unit 204, and a classification unit 205. The storage unit 21 stores classification data 211 and hierarchical information 212.

[0027] The data acquisition unit 201 acquires target data to be classified into one of a plurality of categories in a hierarchical structure. The target data may be any data that can be classified, and may be, for example, text data, image data, or audio data. The target data may be, for example, an item name included in a database or a data table.

[0028] The classification data acquisition unit 202 acquires classification data 211 indicating a plurality of categories into which the target data is to be classified, and identifies candidate categories into which the target data is to be classified. There are no particular limitations on the classification categories as long as they have a hierarchical structure with at least two levels, and a desired category may be specified in advance in the classification data 211.

[0029] The hierarchical structure specifying unit 203 specifies a higher-level category for each of the candidate categories for classification destination based on the hierarchical information 212 indicating the hierarchical structure of the categories. Specifically, the hierarchical information 212 indicates the higher-level category and the lower-level category for each category indicated in the classification destination data 211. Note that only the lower-level category is indicated for the highest-level category, and only the higher-level category is indicated for the lowest-level category. Thus, the hierarchical structure specifying unit 203 can specify the higher-level category for each of the candidate categories for classification destination acquired by the classification destination data acquiring unit 202 by referring to the hierarchical information 212.

[0030] The suitability calculation unit 204 calculates a suitability indicating the degree to which the target data matches the candidate classification destination category identified by the classification destination data acquisition unit 202. The suitability calculation unit 204 also calculates a higher-level suitability indicating the degree to which the target data matches the higher-level category identified by the hierarchical structure identification unit 203. The suitability calculation unit 204 then calculates an overall suitability used by the classification unit 205 for classification from the suitability and higher-level suitability. A method for calculating each of these suitabilities will be described later with reference to FIG. 4.

[0031] The classification unit 205 classifies the target data into one of a plurality of categories based on the overall suitability calculated by the suitability calculation unit 204. As described above, the overall suitability is calculated based on the suitability indicating the degree to which the target data fits into a category and the top suitability, so it can be said that the classification unit 205 classifies the target data based on the suitability indicating the degree to which the target data fits into a category and the top suitability.

[0032] (Calculation method for overall compatibility) A method for calculating the overall suitability will be described with reference to Fig. 4. Fig. 4 is a diagram showing an example of calculating the overall suitability. In this example, the target data is "tap tea", and the candidates for classification are the sub-categories "beer" and "tapioca milk tea".

[0033] In this example, the classification data acquisition unit 202 acquires the minor classification categories "beer" and "tapioca milk tea" as classification candidate categories from the classification data 211. Then, the hierarchical structure identification unit 203 identifies that the higher category of "beer" is "alcohol" and that the higher category of "tapioca milk tea" is "tea". Note that the hierarchical structure identification unit 203 may also identify even higher categories.

[0034] Next, the compatibility calculation unit 204 calculates the compatibility ada(I,C) between the target data and the category, and also calculates the upper compatibility ada between the target data and the upper category. recursive Calculate (I, parent(C)).

[0035] The degree of conformance ada(I,C) between the target data and the category may be an index value that is larger as the possibility of the target data belonging to the category increases. For example, when the target data name and the category name are similar, such as "tap tea" and "tea", it can be said that the target data is highly likely to belong to the category. For this reason, for example, the degree of similarity between the target data name and the category name may be calculated based on the commonality of the character strings contained in those names, and the degree of conformance ada(I,C) may be calculated. In addition, as in an exemplary embodiment 3 described later, the degree of similarity between the target data-related information related to the target data and the category-related information related to the category may be calculated as the degree of conformance ada(I,C).

[0036] In the example of Figure 4, the matching score ada(I,C) of the target data "Tap Tea" to the category "Beer" and the higher matching score ada(I,C) of the target data "Tap Tea" to the category "Alcohol" (a higher category of "Beer") are recursive (I, parent(C)) are both calculated as 0.05. In addition, the matching degree ada(I, C) of the target data "Tapioca Tea" to the category "Tapioca Milk Tea" is calculated as 0.97, and the upper matching degree ada(I, C) of the target data "Tapioca Tea" to the higher category "Tea" is calculated as 0.97. recursive (I, parent(C)) is calculated to be 0.95.

[0037] Here, the compatibility calculation unit 204 calculates the overall compatibility ada using the formula (1) shown in FIG. recursive In addition, α in the formula (1) is a weighting value set between 0 and 1. When the formula (1) is used, when α is less than 0.5, the weighting for the suitability ada(I,C) for the target data category is set to the upper suitability ada(I,C) for the target data category. recursive The weighting is heavier than that for (I, parent(C)). For this reason, it is preferable to set α to less than 0.5. It is also preferable to set the weighting for categories even higher than the higher category even lower than that for the higher category. This allows categories close to the candidate category to be classified into to have a higher influence.

[0038] For example, when α=0.2, the overall compatibility of "Tap tea" with "beer" is ada recursive (I,C)=0.8×0.05+0.2×0.05=0.05. The overall compatibility of "Tapicha" with "Tapioca Milk Tea" is ada recursive (I,C) = 0.8 x 0.97 + 0.2 x 0.95 = 0.97.

[0039] The classification unit 205 classifies the target data based on the overall suitability for each category calculated in this manner. In the example of Fig. 4, the classification unit 205 classifies "Tapioca milk tea" into "Tapioca tea" which has a higher overall suitability.

[0040] As described above, the information processing device 2 according to this exemplary embodiment includes the suitability calculation unit 204 that calculates the overall suitability used for classification by the classification unit 205 from the suitability for the category of the target data and the top suitability. In calculating the overall suitability, the suitability calculation unit 204 may weight the suitability more heavily than the top suitability.

[0041] According to the above configuration, the overall suitability is calculated by weighting the suitability of the target data for a certain category more heavily than the suitability of the category's upper category, and the classification unit 205 classifies the target data using this overall suitability. Therefore, according to the information processing device 2 according to this exemplary embodiment, in addition to the effects of the information processing device 1 according to the exemplary embodiment 1, the effect of being able to classify the target data by placing importance on the suitability of the target data for the category can be obtained.

[0042] (Classification method flow) The flow of the classification method according to this exemplary embodiment will be described with reference to Fig. 5. Fig. 5 is a diagram for explaining the classification method executed by the information processing device 2. Fig. 5 also shows an example of a hierarchical structure of target data and categories into which the target data is classified, together with a flowchart showing the flow of the classification method.

[0043] In S21, the data acquisition unit 201 acquires target data to be classified into one of a plurality of categories in a hierarchical structure. For example, as shown in the figure, if the target data is the word "Tap tea" (product name), the data acquisition unit 201 acquires text data of "Tap tea" as the target data.

[0044] In S22, the classification destination data acquisition unit 202 acquires the classification destination data 211 stored in the storage unit 21, and identifies candidate categories for the classification destination of the target data acquired in S21. For example, in the example of Fig. 5, when classifying the target data "Tapioca tea" into a minor category, the classification destination data acquisition unit 202 identifies the minor categories "beer", "shochu", "tapioca milk tea", and "green tea" from among the major to minor categories shown in the classification destination data 211.

[0045] In S23, the hierarchical structure specification unit 203 specifies a higher-level category of the category specified in S22 based on the hierarchical information 212. If there is a higher-level category above the specified higher-level category, the hierarchical structure specification unit 203 may also specify the higher-level category. This process may be repeated until the top-level category is specified. For example, in a case where three hierarchical levels of categories, namely, major classification, medium classification, and minor classification, are specified, when a minor classification category is specified in S22, the hierarchical structure specification unit 203 may specify at least a medium classification category and further specify a major classification category. If there is no higher-level category above the category specified in S22, classification may be performed based on the similarity between the object-related information and the category-related information.

[0046] For example, when "beer" is identified in S22 from among the categories shown in Fig. 5, the hierarchical structure identification unit 203 needs to identify at least "alcohol" as a higher-level category, and may also identify "drinks", which is a higher-level category of "alcohol". The same applies to other categories.

[0047] In S24, the compatibility calculation unit 204 selects one of the multiple categories identified in S22. Then, in the following S25, the compatibility calculation unit 204 calculates the compatibility of the target data acquired in S21 with each of the categories identified in S22. The compatibility calculation unit 204 also calculates the higher compatibility of the target data acquired in S21 with each of the higher categories identified in S23. Then, in S26, the compatibility calculation unit 204 calculates an overall compatibility from each compatibility calculated in S25.

[0048] In S27, the compatibility calculation unit 204 judges whether or not the calculation of the overall compatibility has been completed for all of the multiple categories identified in S22. If it is judged that the calculation has been completed (YES in S27), the compatibility calculation unit 204 proceeds to the process of S28. On the other hand, if it is judged that the calculation has not been completed (NO in S27), the compatibility calculation unit 204 returns to the process of S24 and selects one category that has not been used in the calculation of the overall compatibility.

[0049] For example, assume that in S22, four minor categories are identified from among the categories shown in Fig. 5, and in S23, medium and major categories are identified for each category. In this case, in the processes of S24 to S27, the suitability of "Tap Tea" to the category "beer" is calculated, and the higher suitability of each of the higher categories of "Tap Tea", "alcohol" and "drinks", is calculated, and the overall suitability of "Tap Tea" to the category "beer" is calculated from these suitability and higher suitability. In addition, the overall suitability of the remaining three minor categories, "shochu", "tapioca milk tea", and "green tea", is calculated in the same manner.

[0050] In S28, the classification unit 205 classifies the target data into the category with the highest overall suitability among the multiple categories identified in S22, thereby completing the classification method of FIG.

[0051] Exemplary embodiment 3 The third exemplary embodiment of the present invention will be described in detail with reference to the drawings. Note that components having the same functions as those described in the second exemplary embodiment are denoted by the same reference numerals, and the description thereof will not be repeated.

[0052] (Configuration of information processing device 2A) The configuration of an information processing device 2A according to this exemplary embodiment will be described with reference to Fig. 6. Fig. 6 is a block diagram showing the configuration of the information processing device 2A. The information processing device 2A differs from the information processing device 2 shown in Fig. 3 in that it includes a related information acquisition unit 204A and that the compatibility calculation unit 204 is replaced with a compatibility calculation unit 204B.

[0053] The related information acquisition unit 204A acquires target related information related to the target data, category related information related to a candidate category into which the target data is to be classified, and higher category related information related to a higher category than the category.

[0054] The target related information may be any information related to the target data. Similarly, the category related information may be any information related to a category, and the higher category related information may be any information related to a higher category. In the following, the target related information, category related information, and higher category related information may be collectively referred to simply as related information. A method for acquiring such related information will be described later.

[0055] The suitability calculation unit 204B calculates the suitability of the target data to the category based on the degree of similarity between the target related information and the category related information. The suitability calculation unit 204B also calculates the higher-level suitability based on the degree of similarity between the target related information and the higher-level category related information.

[0056] As described above, the information processing device 2A according to this exemplary embodiment calculates a compatibility degree based on the degree of similarity between target-related information related to the target data and category-related information related to the category, and is equipped with a compatibility calculation unit 204B that calculates a higher compatibility degree based on the degree of similarity between the target-related information and higher-level category-related information related to the higher-level category.

[0057] When the object-related information related to the object data is similar to the category-related information related to the category, the object data is likely to fit the category. Similarly, when the object-related information is similar to the superordinate category-related information related to the superordinate category, the object data is likely to fit the superordinate category.

[0058] Therefore, according to the above configuration in which the target data is classified based on the compatibility calculated based on the degree of similarity between the target-related information and the category-related information, and the higher compatibility calculated based on the degree of similarity between the target-related information and the higher-level category-related information, the effect of being able to classify the target data into appropriate categories is obtained.

[0059] (How to obtain related information) The related information acquisition unit 204A may acquire related information by searching. The search location may be a location where various information that may be related to the target data is recorded. For example, the related information acquisition unit 204A may search for information related to the target data in a related information DB (database) that records various information that may be related to the target data, and acquire information detected by this search as the target related information. The related information DB may be stored in the storage unit 21, may be stored in a device external to the information processing device 2A, or may be provided in a distributed manner in multiple locations.

[0060] For example, if the target data is text data indicating the names of products, a related information DB may be used that records various text data such as descriptions of various products, reviews of various products, etc. In addition, for example, a database or a data lake of a company that handles products or services related to the target data may be used as the related information DB.

[0061] Also, for example, a database that stores various data extracted by data enrichment targeting various data related to various products and services that may be related to the target data may be used as the related information DB. Data enrichment is a service that enhances the utility value of the target data by extracting various information related to the target data and using it as additional information for the data. In this case, the category determined by the information processing device 2A may be added to the related information DB as information related to the target data. In this case, it can be said that the information processing device 2A performs data enrichment of the target data.

[0062] Furthermore, when the target data is image data, the related information acquiring unit 204A may search the related information DB for images similar to the target data or text data related to the target data.

[0063] Also, for example, the related information acquisition unit 204A may acquire search results obtained by performing a web search on the target data as the target related information. In either search, the search method is not particularly limited, and for example, in the case of a search using text data, a full-sentence match search or a partial-sentence match search may be performed.

[0064] Similarly to the target related information, the related information acquiring unit 204A acquires category related information related to the category of the candidate for classification destination, and acquires higher-level category related information related to the category higher than the category.

[0065] The related information may be acquired by the same method or different methods. Also, each piece of information acquired by different acquisition methods may be regarded as related information. For example, the related information acquisition unit 204A may acquire, as the target related information, both a search result obtained by performing a web search on the target data and a search result obtained by performing a search in a related information DB on the target data.

[0066] (Method of calculating similarity between related information and method of calculating compatibility based on similarity) The compatibility calculation unit 204B calculates the similarity between each piece of related information acquired as described above, and calculates the compatibility between the target data and the category based on the similarity. Here, a method for calculating the similarity between related information and a method for calculating the compatibility based on the similarity will be described with reference to Fig. 7. Fig. 7 is a diagram showing an example of calculating the compatibility based on the similarity between web pages detected in a web search.

[0067] Figure 7 shows the webpage P detected as the top search result for the target data "Tap tea". I 1 and the web page P found as the second top search result. I FIG. 7 shows the top search result for the category "tea", web page P C 1 and the second top result, webpage P C It shows 2.

[0068] In this way, by performing a web search on the target data "Tapioca", various web pages containing the character string "Tapioca" are detected. Similarly, by performing a web search on the category "tea", various web pages containing the character string "tea" are detected. Note that the related information acquisition unit 204A does not need to treat all web search results as related information, and only needs to acquire search results necessary for calculating the similarity as related information. For example, the related information acquisition unit 204A may acquire a predetermined number of the top web search results as related information.

[0069] The matching degree calculation unit 204B calculates the similarity sim(P I i ,P C j ) may be used to calculate the similarity between the search results for the target data "tap tea" and the search results for the category "tea."

[0070] For example, the matching degree calculation unit 204B calculates the degree of overlap of words used in the web pages or documents to be subjected to similarity calculation, the degree of overlap of domain names, or the degree of overlap of words included in file paths as the similarity between the web pages sim(P I i ,P C j For example, the overlapping degree may be calculated by the Jaccard-Index. In this case, the similarity between web pages is calculated as sim(P I i ,P C j ) is expressed by the following formula: sim(P I i ,P C j )=J(bow(P I i ),bow(P C j )) In addition, bow(P I i ) is a web page P Ii Similarly, bow(P C j ) is a web page P C j Of course, the Jaccard-Index is just one example, and any method for calculating the similarity between sets obtained from each search result can be applied.

[0071] The compatibility calculation unit 204B may use the similarity between each web page calculated as described above to calculate the compatibility between the target data and the category by the formula (2) shown in Fig. 7. In formula (2), r(i, j) is a weight. In other words, when formula (2) is used, the compatibility calculation unit 204B performs a calculation of multiplying the similarity between the web pages by the weight r(i, j) corresponding to the search rank for all combinations of each search rank from 1st to 10th, and calculates the sum of each calculation result as the compatibility between the target data and the category.

[0072] Of course, it is not essential to multiply by a weight. However, multiplying by a weight is preferable because it makes it possible to increase the likelihood that a valid similarity is calculated. For example, the weighting for the degree of similarity between top search results may be heavier than the weighting for the degree of similarity between lower search results. This is because top search results are often more closely related to the target data or category than lower search results. Specifically, for example, r(i,j)=(1 / i)·(1 / j) may be used.

[0073] The method of calculating the similarity shown in Fig. 7 can also be applied to calculating the similarity of search results from a search of a related information DB. Also, the compatibility between the target data and a higher-level category, i.e., the higher-level compatibility, can be calculated in the same manner as above.

[0074] Here, when a search is performed on the web or a related information DB, a variety of search results may be obtained, ranging from those highly related to the target data or category to those less related. Therefore, if the search results contained in the target related information and the category related information are all less related to the target data or category, it is possible that an appropriate similarity will not be calculated.

[0075] Therefore, in the information processing device 2A according to this exemplary embodiment, as described above, target-related information indicating the top to bottom search results obtained by searching target data may be used. Also, category-related information indicating the top to bottom search results obtained by searching categories may be used. Furthermore, higher-level category-related information indicating the top to bottom search results obtained by searching higher-level categories may be used. In this case, the compatibility calculation unit 204B calculates the compatibility based on the degree of similarity of each search result from the top to the bottom indicated by each piece of related information.

[0076] According to this configuration, in addition to the effects of the information processing device 1 according to the exemplary embodiment 1, the effect of increasing the probability that each piece of related information contains search results that are highly related to the target data, the category, and the higher-level category, thereby increasing the accuracy of the degree of matching, can be obtained. Even if each piece of related information contains search results that are not highly related to the target data, the category, and the higher-level category, it becomes possible to calculate an appropriate degree of matching overall.

[0077] Furthermore, as described above, in the information processing device 2A according to this exemplary embodiment, the compatibility calculation unit 204B may be configured to weight the degree of similarity between higher-ranked search results more heavily than the degree of similarity between lower-ranked search results when calculating the compatibility.

[0078] In many cases, the top search results are more closely related to the target data or category than the bottom search results. Therefore, according to the information processing device 2A according to this exemplary embodiment, in addition to the effect of the information processing device 1 according to the exemplary embodiment 1, the effect of being able to increase the accuracy of calculating a valid degree of matching can be obtained.

[0079] (Classification method flow) The flow of the classification method according to this exemplary embodiment will be described with reference to Fig. 8. Fig. 8 is a flow diagram showing the flow of the classification method executed by the information processing device 2A. Note that S31 to S33 and S39 to S41 are similar to S21 to S23 and S27 to S29 in Fig. 5, and therefore the description will not be repeated here.

[0080] In S34, the related information acquisition unit 204A acquires target related information that is related information of the target data acquired in S31. For example, the related information acquisition unit 204A may acquire a top predetermined number of search results of the search for the target data as the target related information.

[0081] In S35, the related information acquisition unit 204A selects one of the multiple categories identified in S32. Then, in the following S36, the related information acquisition unit 204A acquires information related to the category selected in S35 as category related information. For example, the related information acquisition unit 204A may acquire a predetermined number of top results from the search results for the category as the category related information, similar to S34.

[0082] In S37, the related information acquisition unit 204A acquires information related to the upper category (specified in S33) of the category selected in S35 as upper category related information. For example, the related information acquisition unit 204A may acquire a predetermined number of top search results of the search for the upper category as upper category related information, similarly to S34 and S36.

[0083] In S38, the suitability calculation unit 204B calculates the suitability of the target data to the category based on the similarity between the object-related information acquired in S34 and the category-related information acquired in S36. The suitability calculation unit 204B also calculates the upper suitability of the target data to the upper category based on the similarity between the object-related information acquired in S34 and the upper category-related information acquired in S37. Then, in S39, the suitability calculation unit 204B calculates an overall suitability from the suitability calculated in S38 and the upper suitability.

[0084] [Modifications] The execution entity of each process described in each of the above exemplary embodiments is arbitrary and is not limited to the above examples. In other words, an information processing system having the same functions as the information processing devices 1, 2, and 2A can be constructed by a plurality of devices that can communicate with each other. For example, an information processing system having the same functions as the information processing devices 2 and 2A can be constructed by distributing the blocks shown in Figures 3 and 6 among a plurality of devices.

[0085] [Software implementation example] Some or all of the functions of the information processing devices 1, 2, and 2A may be realized by hardware such as an integrated circuit (IC chip), or may be realized by software.

[0086] In the latter case, the information processing devices 1, 2, and 2A are realized, for example, by a computer that executes instructions of a program, which is software that realizes each function. An example of such a computer (hereinafter, referred to as computer C) is shown in FIG. 9. The computer C includes at least one processor C1 and at least one memory C2. The memory C2 stores a program P for operating the computer C as the information processing devices 1, 2, and 2A. In the computer C, the processor C1 reads and executes the program P from the memory C2, thereby realizing each function of the information processing devices 1, 2, and 2A.

[0087] The processor C1 may be, for example, a central processing unit (CPU), a graphic processing unit (GPU), a digital signal processor (DSP), a micro processing unit (MPU), a floating point number processing unit (FPU), a physics processing unit (PPU), a microcontroller, or a combination of these. The memory C2 may be, for example, a flash memory, a hard disk drive (HDD), a solid state drive (SSD), or a combination of these.

[0088] The computer C may further include a RAM (Random Access Memory) for expanding the program P during execution and for temporarily storing various data. The computer C may further include a communication interface for transmitting and receiving data to and from other devices. The computer C may further include an input / output interface for connecting input / output devices such as a keyboard, a mouse, a display, and a printer.

[0089] Furthermore, the program P can be recorded on a non-transitory tangible recording medium M that can be read by the computer C. Such a recording medium M can be, for example, a tape, a disk, a card, a semiconductor memory, or a programmable logic circuit. The computer C can obtain the program P via such a recording medium M. Furthermore, the program P can be transmitted via a transmission medium. Such a transmission medium can be, for example, a communication network or broadcast waves. The computer C can also obtain the program P via such a transmission medium.

[0090] [Additional Note 1] The present invention is not limited to the above-described embodiment, and various modifications are possible within the scope of the claims. For example, embodiments obtained by appropriately combining the technical means disclosed in the above-described embodiment are also included in the technical scope of the present invention.

[0091] [Additional Note 2] Some or all of the above-described embodiments can be described as follows. However, the present invention is not limited to the following described aspects.

[0092] (Appendix 1) An information processing device comprising: a data acquiring means for acquiring target data to be classified into one of a plurality of categories in a hierarchical structure, and a classifying means for classifying the target data into one of the plurality of categories based on a suitability indicating the degree to which the target data fits into the category and a higher suitability indicating the degree to which the target data fits into a higher category of the category. With this configuration, the target data can be automatically classified without using a classifier constructed by machine learning.

[0093] (Appendix 2) The information processing device according to claim 1, further comprising a fitness calculation means for calculating an overall fitness used by the classification means for the classification from the fitness and the top fitness, wherein the fitness calculation means weights the fitness more heavily than the top fitness in calculating the overall fitness. With this configuration, it is possible to classify target data by placing importance on the fitness for the category of the target data.

[0094] (Appendix 3) The information processing device according to claim 1 or 2, further comprising a compatibility calculation means for calculating the compatibility based on a degree of similarity between object-related information related to the object data and category-related information related to the category, and calculating the higher compatibility based on a degree of similarity between the object-related information and higher-level category-related information related to the higher-level category. With this configuration, the object data can be classified into appropriate categories.

[0095] (Appendix 4) A classification method including: acquiring target data to be classified into one of a plurality of categories in a hierarchical structure by at least one processor; and classifying the target data into one of the plurality of categories based on a suitability indicating the degree to which the target data fits into the category and a higher suitability indicating the degree to which the target data fits into a higher category of the category. With this configuration, the target data can be automatically classified without using a classifier constructed by machine learning.

[0096] (Appendix 5) A classification program that causes a computer to function as a data acquisition means for acquiring target data to be classified into one of a plurality of categories in a hierarchical structure, and a classification means for classifying the target data into one of the plurality of categories based on a suitability indicating the degree to which the target data fits into the category and a higher suitability indicating the degree to which the target data fits into a higher category of the category. With this configuration, it is possible to automatically classify target data without using a classifier constructed by machine learning.

[0097] [Additional Note 3] A part or all of the above-described embodiments can be further expressed as follows.

[0098] An information processing device comprising at least one processor that executes a process of acquiring target data to be classified into one of a plurality of categories in a hierarchical structure, and a process of classifying the target data into one of the plurality of categories based on a suitability indicating the degree to which the target data fits into the category and a higher-level suitability indicating the degree to which the target data fits into a higher-level category of the category.

[0099] The information processing device may further include a memory, and the memory may store a classification program for causing the processor to execute the process of acquiring the target data and the process of classifying the target data into any one of the multiple categories. The classification program may be recorded in a computer-readable, non-transitory, tangible recording medium. [Explanation of symbols]

[0100] 1. Information processing device 11 Data acquisition unit (data acquisition means) 12 Classification section (classification means) 2. Information processing equipment 201 Data acquisition unit (data acquisition means) 204 Compatibility calculation unit (compatibility calculation means) 205 Classification section (classification means) 2A Information processing equipment 204B Compatibility Calculation Unit (Fitness Calculation Means)

Claims

1. A data acquisition means for acquiring target data to be classified into any one of a plurality of categories in a hierarchical structure; a classification means for classifying the target data into one of the multiple categories based on a degree of suitability indicating the degree to which the target data matches the category and a higher degree of suitability indicating the degree to which the target data matches a higher category than the category; a suitability calculation means for calculating the suitability based on a degree of similarity between object-related information showing the top search results to the bottom search results obtained by searching the object data and category-related information showing the top search results to the bottom search results obtained by searching the category, and for calculating the top suitability based on a degree of similarity between the object-related information and top category-related information showing the top search results to the bottom search results obtained by searching the top category, The compatibility calculation means calculates the compatibility based on the degree of similarity between each search result from the top to the bottom shown in the object-related information and the category-related information, and calculates the higher-level compatibility based on the degree of similarity between each search result from the top to the bottom shown in the object-related information and the higher-level category-related information.

2. The information processing device of claim 1, wherein the compatibility calculation means calculates an overall compatibility used by the classification means for the classification from the compatibility and the top compatibility, and in calculating the overall compatibility, the weight of the compatibility is made heavier than the weight of the top compatibility.

3. At least one processor Obtaining target data to be classified into one of a plurality of categories in a hierarchical structure; classifying the target data into any one of the multiple categories based on a degree of suitability indicating the degree to which the target data matches the category and a higher degree of suitability indicating the degree to which the target data matches a higher category of the category; calculating the degree of suitability based on a degree of similarity between object-related information indicating top search results to bottom search results obtained by searching the object data and category-related information indicating top search results to bottom search results obtained by searching the category; calculating the degree of top-level compatibility based on a degree of similarity between the object-related information and top-level category-related information indicating top search results to bottom search results obtained by searching the top-level category; In calculating the degree of suitability, the degree of suitability is calculated based on a degree of similarity between each of the search results from top to bottom shown in the object-related information and the category-related information; A classification method, in which the calculation of the higher-level suitability is based on the degree of similarity of each search result from top to bottom indicated in the object-related information and the higher-level category-related information.

4. Computer, a data acquisition means for acquiring target data to be classified into any one of a plurality of categories in a hierarchical structure; a classification means for classifying the target data into one of the categories based on a degree of suitability indicating the degree to which the target data matches the category and a higher degree of suitability indicating the degree to which the target data matches a higher category than the category; and a matching degree calculation means for calculating the matching degree based on a degree of similarity between object-related information showing the top search results to the bottom search results obtained by searching the object data and category-related information showing the top search results to the bottom search results obtained by searching the category, and for calculating the top matching degree based on a degree of similarity between the object-related information and top category-related information showing the top search results to the bottom search results obtained by searching the top category, A classification program, wherein the compatibility calculation means calculates the compatibility based on the degree of similarity of each search result from the top to the bottom shown in the object-related information and the category-related information, and calculates the higher-level compatibility based on the degree of similarity of each search result from the top to the bottom shown in the object-related information and the higher-level category-related information.

Images