Information processing device and information processing method

Abstract

An information processing device according to the present invention is provided with a keyword extraction unit that extracts keywords from each of a plurality of pieces of data to be classified, and a classification unit that classifies the plurality of pieces of data by using related words that are related to extracted keywords, which are the keywords extracted from each of the plurality of pieces of data.

Description

[Technical Field]

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

[0002] There are provided techniques for classifying data groups such as documents (sentences). For example, there is provided a technique for clustering (classifying) sentences using keywords extracted from the sentences (for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2015-060581 Summary of the Invention [Problem to be solved by the invention]

[0004] According to conventional techniques, general terms are extracted from a sentence based on predefined part-of-speech information, and compound words different from general terms are extracted as user terms from a sentence based on the frequency of appearance of the compound words.

[0005] However, conventional techniques do not always allow for appropriate classification of data. For example, conventional techniques classify sentences using keywords extracted from the sentences to be classified. Therefore, when keywords that are different when compared as character strings but conceptually synonymous are extracted, conventional techniques are unable to classify the sentences in the same category. As such, it may be difficult to appropriately classify data simply by using keywords extracted from the data to be classified. Therefore, there is a demand for appropriate classification of data.

[0006] Therefore, the present disclosure proposes an information processing device and an information processing method that can appropriately classify data. [Means for solving the problem]

[0007] In order to solve the above problem, an information processing device according to one embodiment of the present disclosure includes a keyword extraction unit that extracts keywords from each of a plurality of data to be classified, and a classification unit that classifies the plurality of data using related words that are related to the extracted keywords, which are keywords extracted from each of the plurality of data. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 10 is a diagram illustrating an example of a keyword extraction process according to an embodiment of the present disclosure. [Figure 2] FIG. 1 is a diagram illustrating an overview of data classification processing according to an embodiment of the present disclosure. [Figure 3] 10 is a flowchart showing a procedure for executing a classification process. [Figure 4] FIG. 10 is a diagram illustrating an example of data classification. [Figure 5] FIG. 10 is a diagram illustrating an example of a data classification result. [Figure 6] FIG. 10 is a diagram illustrating an example of keyword normalization processing according to an embodiment of the present disclosure. [Figure 7] FIG. 10 is a diagram illustrating an example of data classification using related words. [Figure 8] FIG. 10 is a diagram showing an example of a data classification result using related words. [Figure 9] FIG. 10 is a diagram showing an example of points to be considered when using related words. [Figure 10] FIG. 1 is a diagram illustrating a configuration example of an information processing device according to an embodiment of the present disclosure. [Figure 11] FIG. 2 is a diagram illustrating an example of a classification target data storage unit according to an embodiment of the present disclosure. [Figure 12] 1 is a flowchart illustrating a procedure of information processing according to an embodiment of the present disclosure. [Figure 13] FIG. 10 is a diagram illustrating an example of keyword extraction from data other than text. [Figure 14] FIG. 10 is a diagram illustrating an example of updating a classification. [Figure 15] FIG. 10 is a diagram illustrating an example of an updated classification result. [Figure 16] FIG. 10 is a diagram illustrating a configuration example of an information processing system according to a modified example of the present disclosure. [Figure 17] FIG. 10 is a diagram illustrating an example of classification according to users. [Figure 18] FIG. 2 is a hardware configuration diagram illustrating an example of a computer that realizes the functions of the information processing device. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. Note that the information processing device and information processing method according to the present application are not limited to these embodiments. In addition, in the following embodiments, the same components are designated by the same reference numerals, and redundant description will be omitted.

[0010] The present disclosure will be described in the following order: 1. Embodiment 1-1. Overview of information processing according to an embodiment of the present disclosure 1-1-1. Keyword extraction 1-1-2. Data Classification 1-1-3. Classification processing example 1-1-4. Example of processing using related words 1-1-5. Background and effects 1-2. Configuration of the information processing device according to the embodiment 1-3. Information processing procedure according to the embodiment 1-4. Example of keyword extraction other than text 1-5. Classification update example 1-6.Classification conditions 2. Other embodiments 2-1. Variations 2-1-1. Classification by user 2-2.Other configuration examples 2-3.Other 3. Effects of this disclosure 4. Hardware Configuration

[0011] [1. Embodiment] [1-1. Overview of Information Processing According to an Embodiment of the Present Disclosure] Information processing according to an embodiment of the present disclosure is realized by an information processing device 100 (FIG. 10). The information processing device 100 is a computer that extracts keywords (hereinafter also referred to as "extracted keywords") from each of a plurality of data to be classified, and classifies the plurality of data using keywords (hereinafter also referred to as "associated words") related to the extracted keywords of each of the plurality of data. The keywords here refer to character strings that indicate the meaning (content) of the data. The associated words are broader words of the extracted keywords, keywords that rephrase the extracted keywords (hereinafter also referred to as "paraphrases"), keywords associated with the extracted keywords (hereinafter also referred to as "associated words"), etc., specific examples of which will be described later.

[0012] First, various processes executed by the information processing device 100 will be described below, and the details of the configuration of the information processing device 100 will be described later. Note that, although text data (hereinafter also referred to as "text" or "document") will be mainly described below as an example of data, the multiple data to be classified may also include image data (hereinafter also referred to as "image"). Furthermore, the multiple data to be classified may also include time-series data such as video data (hereinafter also referred to as "video"), audio data (hereinafter also referred to as "audio"), and waveform data (hereinafter also referred to as "waveform"). In other words, the multiple data to be classified may include any type of data as long as it is possible to extract an extracted keyword.

[0013] [1-1-1. Keyword extraction] First, keyword extraction from data will be described using Fig. 1. Fig. 1 is a diagram showing an example of a keyword extraction process according to an embodiment of the present disclosure. The keyword extraction process in Fig. 1 is executed by the keyword extraction unit 152 (see Fig. 10) of the information processing device 100. Specifically, Fig. 1 shows an example of extracting extracted keywords from a sentence X, which is one piece of data DT1, from a data group to be separated.

[0014] The information processing device 100 extracts extracted keywords from a text X using text mining technology (step S1). The information processing device 100 extracts extracted keywords from the text X using morphological analysis, word counting, co-occurrence networks, etc. In FIG. 1, the information processing device 100 extracts character strings such as "dog," "pet," "Chihuahua," "shed," and "cute" from the text X as extracted keywords, as shown in keyword information KX. The information processing device 100 may identify the part of speech of each extracted keyword and associate information on the identified part of speech with each extracted keyword. For example, extracted keywords extracted from the text X are words that have important meanings in the text X.

[0015] The information processing device 100 calculates points for each extracted keyword in the sentence X according to the degree of relevance with the sentence X, and assigns the calculated points to each extracted keyword. For example, the information processing device 100 calculates points for each extracted keyword using the TF-IDF method, which is a method for evaluating the importance of words in a sentence (document). The information processing device 100 also corrects (normalizes) the points assigned to the extracted keywords so that the total sum of the points is 1. In the example of FIG. 1, the information processing device 100 normalizes the points for "dog" to "0.6," "pet" to "0.2," "Chihuahua" to "0.1," "shed" to "0.06," and "cute" to "0.04" for each extracted keyword in the sentence X. As such, since "dog" has the highest point among the extracted keywords in the sentence X, it is highly likely that the sentence X is mainly related to "dogs."

[0016] The information processing device 100 may calculate the points of each extracted keyword using any method capable of calculating points, not limited to TF-IDF. For example, the information processing device 100 may use frequency as the points. In this case, the information processing device 100 may calculate the points such that the higher the frequency of an extracted keyword, the higher the point. Furthermore, when calculating points for data other than text, the information processing device 100 calculates the points according to the type of data. For example, if the data is an image, the information processing device 100 may calculate, for an extracted keyword indicating an object included in the image, the percentage of the object that occupies the image as the point. Furthermore, for example, if the data is a video, the information processing device 100 may calculate, for an extracted keyword indicating an object included in the video, the time the object is visible in the video as the point. Details of data other than text will be described later.

[0017] Furthermore, a limit may be placed on the maximum number of keywords extracted from each data (maximum number of extractions). For example, if the maximum number of extractions is specified as "4," the information processing device 100 extracts extracted keywords from each data, with the upper limit being four. In this case, the information processing device 100 may extract extracted keywords from the data, calculate the points of each extracted keyword, and then exclude all extracted keywords other than the four with the largest points.

[0018] [1-1-2. Data Classification] Next, an overview of data classification based on extracted keywords will be described with reference to Fig. 2. Fig. 2 is a diagram showing an overview of data classification processing according to an embodiment of the present disclosure. The classification processing in Fig. 2 is executed by the classification unit 154 (see Fig. 10) of the information processing device 100.

[0019] The example in Fig. 2 shows an outline of processing when classifying multiple pieces of data such as text A, text B, text C, and text D. Keyword information KA indicates extracted keywords extracted by the information processing device 100 from text A. Keyword information KB, KC, and KD indicate extracted keywords extracted by the information processing device 100 from text B, text C, and text D, respectively.

[0020] The information processing device 100 classifies a plurality of data such as text A, text B, text C, and text D using keyword information KA, KB, KC, KD, etc. (step S2). In the example of Fig. 2, the information processing device 100 classifies a plurality of data such as text A, text B, text C, and text D into a plurality of categories (hereinafter also referred to as "groups") based on the commonality of extracted keywords, as shown in the classification result RS1. The information processing device 100 classifies text A, text B, etc. into a group indicated as "category #1", and classifies text C, text D, etc. into a group indicated as "category #2".

[0021] The information processing device 100 classifies a data group at an appropriate granularity based on the extracted keywords. For example, when there is an existing classification result, the information processing device 100 may use the existing classification result to classify a new classification target. The processing in this case will be described with reference to FIG. 3. FIG. 3 is a flowchart showing the procedure for executing the classification process. For example, when there is a classification result RS1 shown in FIG. 2, a case where new data DX is set as a classification target will be described as an example.

[0022] First, the information processing device 100 determines whether the new data fits into an existing classification (step S11). For example, the information processing device 100 determines whether the data DX fits into either of the existing classifications "Classification #1" or "Classification #2."

[0023] If the new data fits into an existing classification (step S11: Yes), the information processing device 100 determines whether the granularity is appropriate (step S12). For example, if the number of data in the group to which the data DX is added exceeds the upper limit of the number of data belonging to one group (also called the "maximum number of items"), the information processing device 100 determines that the granularity is inappropriate.

[0024] If the granularity is appropriate (step S12: Yes), the information processing device 100 adds the new data to the corresponding classification and ends the process. For example, if the number of data items in "Classification #1" to which the data DX has been added is equal to or less than the maximum number of items, the information processing device 100 adds the data DX to the corresponding "Classification #1" and ends the process.

[0025] If the new data fits into an existing classification (step S11: No), the information processing device 100 executes classification processing (step S13). For example, the information processing device 100 executes classification processing again on the entire data group to which the new data has been added. Furthermore, if the granularity is not appropriate (step S12: No), the information processing device 100 executes classification processing (step S13).

[0026] The above process is an example, and the information processing device 100 uses various information as appropriate to determine whether new data fits into an existing classification or whether the granularity is appropriate. An example of this point will be described below.

[0027] [1-1-3. Classification processing example] An example of the classification process will now be described with reference to Fig. 4 and Fig. 5. Fig. 4 is a diagram showing an example of data classification, and Fig. 5 is a diagram showing an example of the data classification result.

[0028] Here, the granularity of classification has a significant impact on the searchability of a data group. For example, a user of the information processing device 100 sets the granularity of classification by specifying the maximum number of items or the classification threshold. As described above, the maximum number of items is the maximum number of items (data) included in one classification (group). If the number of items (data) in one classification (group) exceeds this value, classification processing is executed. Furthermore, the classification threshold is used as a criterion for determining whether or not to classify, for example, based on whether the score (points) of a group, such as a classification based on a certain common keyword, exceeds a threshold.

[0029] Below, we will explain the specific classification procedure when using the maximum number of items and the classification threshold. Specifically, the maximum number of items is set to "4" and the classification threshold is set to "0.1". The classification threshold is used to compare with the average score of the common keyword for each data in the group classified by the common keyword (hereinafter also referred to as "group score").

[0030] A specific description will now be given using Fig. 4. As shown in the keyword information KA, KB, KC, and KD in Fig. 4, extracted keywords are extracted from four pieces of data, sentence A, sentence B, sentence C, and sentence D, by the information processing device 100, and points for the extracted keywords are calculated. In this case, the data group consists of four pieces of data, sentence A, sentence B, sentence C, and sentence D, which is less than the maximum number of items "4," and the information processing device 100 determines that classification is unnecessary.

[0031] In the case of Fig. 4, the points assigned to the keywords of each data item of sentence A, sentence B, sentence C, and sentence D are normalized so that the sum of the points is 1. For each extracted keyword of sentence A, the information processing device 100 normalizes the points for "tennis" to "0.4", the points for "sports" to "0.3", the points for "professional" to "0.15", the points for "court" to "0.12", and the points for "player" to "0.03". The information processing device 100 similarly normalizes the points for each extracted keyword of sentence B, sentence C, and sentence D.

[0032] Here, a new sentence E is added. As shown in the keyword information KE, the information processing device 100 extracts extracted keywords from sentence E and calculates points for the extracted keywords. For each extracted keyword in sentence E, the information processing device 100 normalizes the points for "hamster" to "0.5", "pet" to "0.4", "birth" to "0.04", "running wheel" to "0.03", and "discipline" to "0.03".

[0033] In this case, the data group consists of five items: sentence A, sentence B, sentence C, sentence D, and sentence E, which exceeds the maximum number of items of "4." Therefore, the information processing device 100 determines that the granularity is inappropriate and performs classification processing (step S21).

[0034] The information processing device 100 performs classification processing using the extracted keywords and the assigned points. The information processing device 100 compares the extracted keywords of each data and identifies data that share the extracted keywords in common. For example, the information processing device 100 identifies that sentence A and sentence B share the keyword "sports" based on the degree of association of each data. The information processing device 100 then calculates that the sum (also referred to as the "total points") of the points "0.3" of the extracted keyword "sports" of sentence A and the points "0.2" of the extracted keyword "sports" of sentence B is "0.5." When sentences A and B are grouped (classified) into a single group "sports," the information processing device 100 calculates that the points of the group "sports," which is the average value (average points) of the extracted keyword "sports" of sentence A and the extracted keyword "sports" of sentence B, is "0.25 (=0.5 / 2)." The point of the group "sports" of "0.25" exceeds the classification threshold of "0.1" which is the classification criterion, and therefore the information processing device 100 determines that the classification of the group "sports" is valid.

[0035] Furthermore, the information processing device 100 finds that sentences C, D, and E share the common extracted keywords "pet" or "training." The information processing device 100 calculates that the total points for grouping sentences C, D, and E under "pet" will be "0.9 (=0.2+0.3+0.4)." The information processing device 100 also calculates that the total points for grouping sentences C, D, and E under "training" will be "0.12 (=0.06+0.03+0.03)." Therefore, the information processing device 100 determines that "pet," which has a larger total point, is more suitable as a keyword for grouping sentences A, B, and E.

[0036] Furthermore, when sentences C, D, and E are grouped under the category of "pets," the average point is "0.3 (=0.9 / 3)," and therefore the information processing device 100 determines that the classification of the group "pets" is valid. Note that when there are multiple common keywords in data corresponding to one group, the information processing device 100 may consider these sentences (documents) to be more similar and add these together for calculation. In the example of FIG. 4, the information processing device 100 determines the points for the group "pets" to be "0.34 (=0.3+0.04)," which is the average point of sentences C, D, and E, "0.3," plus the average point of the "training" points of sentences C, D, and E, "0.04 (=0.12 / 3)."

[0037] As described above, when sentence E is added, the information processing device 100 classifies sentences A and B into the group "sports" and sentences C, D, and E into the group "pets," as shown in classification result RS2.

[0038] Details of the classification result when sentence E is added are shown in Fig. 5. As shown in classification result RS2 in Fig. 5, sentences A and B are classified into group G1 corresponding to the classification word "sports," and sentences C, D, and E are classified into group G2 corresponding to the classification word "pets." In this case, the information processing device 100 may set classification items for bookmarks of the classification word "sports" including sentences A and B and the classification word "pets" including sentences C, D, and E.

[0039] [1-1-4. Example of processing using related words] Next, the classification process using related words will be explained, and explanations of the same points as those in the above-mentioned Figs. 1 to 5 will be omitted as appropriate.

[0040] First, a keyword normalization process including identification of related words will be described with reference to Fig. 6. Fig. 6 is a diagram showing an example of a keyword normalization process according to an embodiment of the present disclosure. The keyword normalization process in Fig. 6 is executed by the keyword normalization unit 153 (see Fig. 10) of the information processing device 100.

[0041] The information processing device 100 uses the extracted keywords to change them to keywords that are more suitable for the classification, to exclude the extracted keywords, or to add keywords.

[0042] First, the exclusion (deletion) of extracted keywords will be described. For example, when extracted keywords are extracted from a text Z related to cooking, keywords such as "delicious" and "make" may be extracted. The information processing device 100 excludes (deletes) such words based on personal concepts or actions (hereinafter also referred to as "exclusion target words") from the extracted keywords. Then, the information processing device 100 performs point normalization again on the extracted keywords after the exclusion target words have been excluded. For example, the information processing device 100 may select keywords corresponding to specific parts of speech, such as adjectives and verbs, as exclusion target words. For example, the information processing device 100 may exclude corresponding extracted keywords using exclusion keyword information indicating the parts of speech to be excluded. Note that the above is just an example, and the information processing device 100 may exclude keywords using any method. For example, the information processing device 100 may use list information of excluded keywords, such as "te," "ni," "o," and "wa," to exclude extracted keywords that match character strings in the list information.

[0043] Next, the identification of related words for an extracted keyword will be described. The information processing device 100 identifies related words for an extracted keyword and associates the related words with the extracted keyword. As described above, related words include superordinate concepts (hypernyms) of keywords, paraphrases, associative words, etc. For example, a superordinate concept of "Chihuahua" is "dog," and a superordinate concept of that is "animal." In other words, "dog" and "animal" can be related words for "Chihuahua." Furthermore, as a paraphrase, a paraphrase of "pet dog" is "pet." In other words, "pet" can be a related word for "pet dog."

[0044] For example, the information processing device 100 identifies related words for each keyword using a concept dictionary that indicates conceptual relationships between languages. For example, various concept dictionaries such as WordNet disclosed in the following documents are used as the concept dictionary. The information processing device 100 identifies related words for each keyword using WordNet. Japanese WordNet<http: / / compling.hss.ntu.edu.sg / wnja / >

[0045] 6 shows an example of extracting extracted keywords from a sentence X, which is one piece of data DT1, from a data group to be separated. The information processing device 100 extracts extracted keywords from the sentence X (step S31). In FIG. 6, the information processing device 100 extracts character strings such as "dog," "pet," "Chihuahua," "shed," and "cute" from the sentence X as extracted keywords, as shown in keyword information KX.

[0046] Then, the information processing device 100 normalizes the extracted keywords extracted from the sentence X (step S32). As shown in the normalized keyword information RKX, the information processing device 100 excludes "cute", which corresponds to the excluded part of speech "adjective", from the extracted keywords. Then, the information processing device 100 corrects (normalizes) the points assigned to each extracted keyword after excluding "cute" so that the total sum of the points is 1. In the example of FIG. 6, the information processing device 100 normalizes the points of "dog" to "0.61", "pet" to "0.21", "Chihuahua" to "0.11", and "shed" to "0.07" for each extracted keyword in the sentence X.

[0047] Furthermore, the information processing device 100 identifies related words for each extracted keyword. The information processing device 100 identifies related words for each of "dog," "Chihuahua," and "shed" using a concept dictionary such as WordNet. In FIG. 6, as shown in the normalized keyword information RKX, the information processing device 100 identifies "pet" and "animal" as related words for "dog," and identifies "animal" as a related word for "pet." The information processing device 100 identifies "dog," "pet," and "animal" as related words for "Chihuahua," and identifies "building" as a related word for "shed."

[0048] Note that the above is just an example, and the information processing device 100 may use any information as long as it can identify related words for each keyword. For example, the information processing device 100 may store a related word list in the storage unit 120 (see FIG. 10 ) that associates each keyword with its related words, and use the related word list to identify related words for each keyword.

[0049] Furthermore, for example, when the information processing device 100 holds personal information of a user, such as a telephone directory or family information, the information processing device 100 may use the user's personal information to identify related words of a keyword. For example, when the information processing device 100 has personal information of user U1 indicating a family relationship, such as the fact that user U1's mother is named "Yuuko," the information processing device 100 may identify the keyword "mother," which is a replacement expression, as a related word of the keyword "Yuuko" for user U1. Furthermore, the information processing device 100 may identify the hypernym "family," which is a superordinate concept of "mother," as a related word of the keyword "Yuuko" for user U1.

[0050] The information processing device 100 assigns the identified related words to the extracted keywords, thereby storing the extracted keywords in association with the identified related words. In this way, the information processing device 100 can perform more conceptual classification by using the identified related words.

[0051] An example of classification processing using related words will now be described with reference to Figures 7 and 8. Figure 7 is a diagram showing an example of data classification using related words. Figure 8 is a diagram showing an example of the results of data classification using related words.

[0052] 7, for the four pieces of data, Sentence A, Sentence B, Sentence C, and Sentence D, related words have been identified for each extracted keyword by the information processing device 100. In this case, the data group consists of four pieces of data, Sentence A, Sentence B, Sentence C, and Sentence D, which is less than the maximum number of items "4," and the information processing device 100 determines that classification is unnecessary.

[0053] Here, a sentence F is newly added. As shown in the normalized keyword information RKE, the information processing device 100 extracts extracted keywords from the sentence F, calculates points for the extracted keywords, and identifies related words for the extracted keywords. For each extracted keyword in the sentence F, the information processing device 100 normalizes the points for "elephant" to "0.6," "zoo" to "0.3," "meal" to "0.05," "expense" to "0.03," and "Ueno" to "0.02." Furthermore, the information processing device 100 identifies "animal" as a related word for "elephant," "meal" and "feed" as related words for "meal," and "Tokyo" and "Japan" as related words for "Ueno."

[0054] The information processing device 100 performs classification processing on the data group of sentences A, B, C, D, and F (step S41). The information processing device 100 performs classification processing using related words. The information processing device 100 compares the extracted keywords and related words of each piece of data to identify data that share the extracted keywords or related words.

[0055] Here, based on the extracted keywords alone, there is no data among sentence A, sentence B, sentence C, and sentence D that has keywords in common with sentence F. On the other hand, when the related words are compared, the common keyword "animal" is present in sentences C, D, and F. Therefore, the information processing device 100 determines that sentences C, D, and F have the extracted keyword "animal" in common.

[0056] 6, the information processing device 100 calculates the average points using the points of the extracted keyword with the highest points among the extracted keywords associated with the related word "animals." The information processing device 100 adds up the points "0.6" of "dog" in sentence C, the points "0.5" of "cat" in sentence D, and the points "0.6" of "elephant" in sentence F, and calculates that the total points for grouping sentences C, D, and F under "animals" is "1.7 (=0.6+0.5+0.6)."

[0057] The above calculation is an example, and points calculated from the extracted keyword points using a predetermined function (hereinafter also referred to as "related word points") may be used when using related words. For example, if the related word is a paraphrase, the points may be 1 times the extracted keyword (i.e., used as is); if the related word is a superordinate concept (hypernym), the points may be 0.8 times the extracted keyword, and then 0.8 times that superordinate concept. The more distant the conceptual relationship, the lower the points may be. This point will be described in more detail later.

[0058] The information processing device 100 determines that the average score when sentences C, D, and E are grouped together as "animals" is "0.57 (=1.7 / 3)", which exceeds the classification threshold of "0.1", and therefore determines that the classification of the group "animals" is valid.

[0059] Furthermore, the information processing device 100 identifies that sentences A and B share the related word "sports." In the example of FIG. 6, the information processing device 100 calculates an average point using the points of the extracted keyword with the highest point value among the extracted keywords associated with the related word "sports." The information processing device 100 adds up the points of the extracted keyword "tennis" of sentence A (0.4) and the points of the extracted keyword "swimming" of sentence B (0.5), and calculates that the total points when sentences A and B are grouped under "sports" is 0.9 (=0.4+0.5). The information processing device 100 determines that the average points when sentences A and B are grouped under "sports" is 0.45 (=0.9 / 2), which exceeds the classification threshold of 0.1, and therefore determines that the classification of the group "sports" is valid.

[0060] As described above, when sentence F is added, the information processing device 100 classifies sentences A and B into the group "sports" and sentences C, D, and F into the group "animals," as shown in the classification result RS11.

[0061] Details of the classification result when sentence F is added are shown in Fig. 8. As shown in the classification result RS11 in Fig. 8, sentences A and B are classified into group G11 corresponding to the classification word "sports," and sentences C, D, and F are classified into group G12 corresponding to the classification word "animals." In this case, the information processing device 100 may set classification items for bookmarks of the classification word "sports" including sentences A and B and the classification word "animals" including sentences C, D, and F.

[0062] As described above, the information processing device 100 classifies data using words related to extracted keywords extracted from each piece of data. This allows the information processing device 100 to appropriately classify data that cannot be classified into one group when only the extracted keywords are used, such as sentences C, D, and F, into one group. Therefore, the information processing device 100 can appropriately classify data.

[0063] Note that, since the extracted keywords of sentence A and sentence B include "sports," the points of the group may also be calculated taking these points into consideration. For example, the information processing device 100 may calculate the points of the group "sports" using the point "0.7 (=0.4+0.3)" of the keyword "sports" of sentence A and the point "0.7 (=0.5+0.2)" of the keyword "sports" of sentence B. In this case, the information processing device 100 may calculate the points of the group "sports" as "0.7 (=1.4 / 2)" using the total point "1.4 (=0.7+0.7)" obtained by adding up the point "0.7" of the keyword "sports" of sentence A and the point "0.7" of the keyword "sports" of sentence B.

[0064] The information processing device 100 may classify data that has common extracted keywords and related words into one group. For example, the information processing device 100 may classify data that has "travel" only in the extracted keywords and data that has "travel" only in the related words into one group, "travel." In this way, the information processing device 100 may perform classification processing based on the commonality of the entire keywords between the extracted keywords and the related words.

[0065] An example of the related word points mentioned above will now be described with reference to Fig. 9. Fig. 9 is a diagram showing an example of points when related words are used. Specifically, Fig. 9 shows an example of the case where related word points are applied to the group G11 corresponding to the classification word "sports" in the classification result RS11 of Fig. 8.

[0066] In the example of FIG. 9, the information processing device 100 calculates the related word points for the related word "sports" of the sentence A by multiplying the point "0.4" of the extracted keyword "tennis" by the coefficient "0.8" because the related word "sports" is a hypernym of the extracted keyword "tennis." Specifically, as shown in the normalized keyword information RKA2, the information processing device 100 calculates the related word points for the related word "sports" of the sentence A as "0.32 (=0.4*0.8)." Furthermore, as shown in the normalized keyword information RKB2, the information processing device 100 calculates the related word points for the related word "sports" of the sentence B as "0.4 (=0.5*0.8)." In this case, the information processing device 100 calculates that the average points for the sentences A and B grouped together as "sports" will be "0.36 (=(0.32+0.4) / 2)."

[0067] [1-1-5. Background and effects] For example, it is difficult to determine the content of various data groups at a glance, such as time-series data such as emails, bookmarks, images, and videos, and the searchability becomes worse as the amount of data in the data group increases.Therefore, it is desirable to analyze the content of each data in the data group and classify it semi-automatically by dividing it into folders and assigning tags.

[0068] For example, there is a technology that suggests which folder new emails should be sorted into by setting related keywords to email folders in advance. However, the keywords must first be manually assigned, and the more email folders there are, the more difficult it becomes to establish rules to correctly classify data with common elements. In addition, if the number of emails in a single email folder increases and searchability deteriorates, it becomes necessary to reset the sorting method to make it more searchable.

[0069] In addition, there are an increasing number of cases where documents are photographed and managed as digital image data. However, in recent years, image classification has often only been limited to classification as documents. Therefore, a method is needed to analyze the documents in the images and classify their contents.

[0070] Meanwhile, the information processing device 100 extracts keywords from various types of data, such as time-series data (e.g., text, images, and videos), and classifies the data into groups using related words related to the extracted keywords. In this way, the information processing device 100 sorts data according to its content from a large amount of data, regardless of the data type, making it easier to find the desired data. Furthermore, the information processing device 100 eliminates the need for the user to manually configure sorting settings to collect the desired data, thereby reducing the occurrence of human error. The information processing device 100 also helps identify documents with the same content, making document organization easier. The information processing device 100 can also classify the content of images containing character strings. This allows the information processing device 100 to collectively manage document groups and image groups containing character strings according to the same classification rules.

[0071] [1-2. Configuration of the information processing device according to the embodiment] Next, a configuration of an information processing device 100, which is an example of an information processing device that executes information processing according to an embodiment, will be described. Fig. 10 is a diagram showing an example of the configuration of the information processing device 100 according to an embodiment of the present disclosure. For example, the information processing device 100 shown in Fig. 10 is an example of an information processing device. The information processing device 100 is a computer that realizes functions as an information processing device, which will be described later.

[0072] 10, the information processing device 100 includes a communication unit 11, an input unit 12, a display unit 13, a storage unit 14, and a control unit 15. In the example of Fig. 10, the information processing device 100 includes the input unit 12 (e.g., a keyboard, a mouse, etc.) that accepts various operations from an administrator of the information processing device 100, and the display unit 13 (e.g., a liquid crystal display, etc.) that displays various information.

[0073] The communication unit 11 is realized by, for example, a NIC (Network Interface Card), a communication circuit, etc. The communication unit 11 is connected to a communication network N (a network such as the Internet) by wire or wirelessly, and transmits and receives information to and from other devices, etc. via the communication network N.

[0074] The input unit 12 receives various operations input from an administrator or the like (also referred to as an "operator") of the information processing device 100. The input unit 12 accepts inputs from the operator. The input unit 12 may accept specifications from the operator regarding classification conditions such as the maximum number of items and the maximum number of categories. The input unit 12 may accept various operations from the operator via a keyboard, mouse, or touch panel provided in the information processing device 100.

[0075] The display unit 13 displays various types of information. The display unit 13 is a display device (display unit) such as a display, and displays various types of information. The display unit 13 displays information on each piece of data such as extracted keywords and related words. The display unit 13 displays information on the classification results by the classification unit 154.

[0076] Furthermore, the information processing device 100 may have a functional configuration for outputting information, not limited to the display unit 13. The information processing device 100 may have a function for outputting information as sound. For example, the information processing device 100 may have an audio output unit such as a speaker that outputs sound.

[0077] The storage unit 14 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 14 has a concept dictionary storage unit 141, a classification target data storage unit 142, and a classification condition storage unit 143.

[0078] Although not shown in the figure, various information related to a concept dictionary that indicates conceptual relationships between languages ​​is stored in the concept dictionary storage unit 141. For example, the concept dictionary storage unit 141 stores information from WordNet.

[0079] A concept dictionary for each language may be stored in the concept dictionary storage unit 141. In this case, the concept dictionary storage unit 141 may store concept dictionaries for various languages, such as a Japanese concept dictionary, an English concept dictionary, a German concept dictionary, and a Chinese concept dictionary.

[0080] The classification target data storage unit 142 according to the embodiment stores information about data to be classified. For example, the classification target data storage unit 142 stores information such as data, extracted keywords, points, and related words corresponding to the data. FIG. 11 is a diagram illustrating an example of a classification target data storage unit according to an embodiment of the present disclosure. FIG. 11 illustrates an example of the classification target data storage unit 142 according to the embodiment. In the example illustrated in FIG. 11, the classification target data storage unit 142 includes items such as "data," "type," "extracted keywords," "points," and "related words."

[0081] "Data" indicates each piece of data. A data ID that identifies each piece of data may be stored in association with each piece of data. "Type" indicates the type of each piece of data. "Type" stores information indicating the type of the data, such as text, image, video, audio, waveform, other time-series data, etc.

[0082] "Extracted keyword" indicates the extracted keyword extracted from the data. "Point" indicates the point (score) that indicates the degree of relevance of the extracted keyword to the data from which it was extracted. "Related word" indicates the related word of the extracted keyword. As shown in Figure 11, multiple related words may be associated with one extracted keyword.

[0083] 11, the example shows that keywords such as "tennis," "sports," "professional," "court," and "player" have been extracted from sentence A, which is data of the type "sentence." The example also shows that the extracted keyword "tennis" is associated with the hypernym "sports" as a related word.

[0084] Furthermore, the video MV, which is data of the type "video," indicates that keywords such as "dog," "park," "Chihuahua," and "Frisbee" have been extracted as extracted keywords. The extracted keyword "dog" indicates that the hypernym "animal" is associated as a related word. The extracted keyword "Chihuahua" indicates that the hypernym "dog" and the further hypernym "animal" are associated as related words. The extracted keyword "Frisbee" indicates that the hypernym "toy" is associated as a related word. Related words are not limited to hypernyms, but may also be paraphrases of the extracted keywords or associative words associated with the extracted keywords.

[0085] The classification target data storage unit 142 may store various types of information depending on the purpose, without being limited to the above. For example, the classification target data storage unit 142 may store information indicating the classification result of each piece of data. The classification target data storage unit 142 may store information indicating the group to which each piece of data belongs.

[0086] The classification condition storage unit 143 stores various information related to the conditions for performing classification. For example, the classification condition storage unit 143 stores information on the maximum number of items indicating the upper limit of the number of data items belonging to each group. The classification condition storage unit 143 also stores information on a classification threshold, which is a predetermined value used to determine whether the points of each group satisfy a standard. The classification condition storage unit 143 also stores information on the maximum number of categories, which is a category number threshold indicating the upper limit of the number of groups to be classified. The classification condition storage unit 143 also stores information (such as a coefficient such as 0.8) for calculating points used when grouping using related words. For example, the classification condition storage unit 143 also stores information on the number of processing times, which indicates the upper limit of the number of repetitions of the classification process until the classification is finalized. For example, the classification condition storage unit 143 stores exclusion keyword information indicating keywords to be excluded.

[0087] 10, the description will be continued. The control unit 15 is realized, for example, by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like executing a program (for example, an information processing program according to the present disclosure) stored inside the information processing device 100 using a RAM (Random Access Memory) or the like as a working area. The control unit 15 is a controller, and may be realized, for example, by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

[0088] 10, control unit 15 has an acquisition unit 151, a keyword extraction unit 152, a keyword normalization unit 153, a classification unit 154, and a transmission unit 155, and realizes or executes the functions and actions of information processing described below. Note that the internal configuration of control unit 15 is not limited to the configuration shown in FIG. 10, and may be any other configuration as long as it performs the information processing described below.

[0089] The acquisition unit 151 acquires various types of information. The acquisition unit 151 acquires various types of information from the storage unit 14. The acquisition unit 151 acquires information accepted by the input unit 12. The acquisition unit 151 acquires various types of information from the storage unit 14. The acquisition unit 151 acquires various types of information from the concept dictionary storage unit 141, the classification target data storage unit 142, and the classification condition storage unit 143. The acquisition unit 151 functions as a receiving unit that receives various types of information from an external information processing device. The acquisition unit 151 receives various types of information from a terminal device used by a user (such as the user terminal 10 in FIG. 16).

[0090] The keyword extraction unit 152 performs processing to extract keywords. The keyword extraction unit 152 extracts keywords from data in order to classify a data group. The keyword extraction unit 152 extracts keywords based on information stored in the storage unit 14. The keyword extraction unit 152 extracts keywords from each piece of data stored in the data to be categorized storage unit 142 based on information. The keyword extraction unit 152 stores the extracted keywords extracted from the data in the data to be categorized storage unit 142 in association with the data from which they were extracted.

[0091] The keyword extraction unit 152 extracts keywords from each of a plurality of data to be classified. The keyword extraction unit 152 extracts keywords from each of a plurality of data including text data, image data, or time-series data. The keyword extraction unit 152 extracts keywords from each of a plurality of data including video data, audio data, or waveform data. The keyword extraction unit 152 performs the keyword extraction process shown in Figs. 1 and 6.

[0092] The keyword extraction unit 152 functions as a data analysis unit that extracts extracted keywords from data other than text. The keyword extraction unit 152 extracts extracted keywords from images using image analysis technology. For example, the keyword extraction unit 152 analyzes an image, recognizes objects contained in the image, and extracts the names of the recognized objects as extracted keywords. The keyword extraction unit 152 also analyzes the image, estimates the proportion (occupancy) of the object contained in the image that occupies the image, and calculates the points of the extracted keywords that indicate the objects based on the estimated occupancy of the objects.

[0093] Furthermore, the keyword extraction unit 152 recognizes characters in an image or video. The keyword extraction unit 152 uses character recognition technology to recognize characters from an image or video in which characters are captured. When the image is an image containing a character string, the keyword extraction unit 152 may extract a character string from the image containing the character string using a character recognizer or the like, and extract an extracted keyword from the extracted character string. The keyword extraction unit 152 may have the function of a character recognizer. When the image is an image containing a character string, the keyword extraction unit 152 extracts extracted keywords from the character string extracted from the image, in the same way as when the data is text, and calculates points for each extracted keyword.

[0094] The keyword extraction unit 152 extracts extracted keywords from time-series data using technology related to time-series data analysis. The keyword extraction unit 152 extracts extracted keywords from videos using technology related to video analysis. For example, the keyword extraction unit 152 analyzes the video, recognizes objects included in the video, and extracts the names of the recognized objects as extracted keywords. The keyword extraction unit 152 also analyzes the video, identifies the time when the object appears in the video, and calculates the points of the extracted keywords that indicate the object based on the identified time.

[0095] The keyword extraction unit 152 uses a technology related to voice analysis to extract extracted keywords from the voice. For example, the keyword extraction unit 152 analyzes the voice and converts the voice into text (character information). Furthermore, the keyword extraction unit 152 extracts extracted keywords from the text converted from the voice, just as in the case where the data is text, and calculates points for each extracted keyword.

[0096] Furthermore, the keyword extraction unit 152 may extract extracted keywords from the waveform using a technique related to waveform analysis. Note that the above is just one example, and the keyword extraction unit 152 may extract extracted keywords and calculate points using any method as long as it can extract extracted keywords from data and calculate points. The keyword extraction unit 152 extracts extracted keywords from data using a method appropriate for the type of data, and calculates points for the extracted keywords.

[0097] The keyword normalization unit 153 performs processing to normalize keywords of each data. The keyword normalization unit 153 normalizes keywords extracted from data. The keyword normalization unit 153 assigns related words to the extracted keywords of the data. The keyword normalization unit 153 identifies related words of the extracted keywords, associates them with the data from which the extracted keywords were extracted, and stores them in the classification target data storage unit 142. The keyword normalization unit 153 identifies related words of the extracted keywords using the concept dictionary stored in the concept dictionary storage unit 141.

[0098] The keyword normalization unit 153 excludes keywords that satisfy the exclusion conditions from the extracted keywords. The keyword normalization unit 153 excludes keywords that correspond to parts of speech to be excluded from the extracted keywords. The keyword normalization unit 153 excludes keywords that satisfy the keyword exclusion conditions stored in the classification condition storage unit 143 from the extracted keywords. The keyword normalization unit 153 performs keyword normalization processing shown in FIG. 6.

[0099] The classification unit 154 performs a process of classifying data. The classification unit 154 classifies a plurality of pieces of data stored in the classification target data storage unit 142. The classification unit 154 determines whether the newly classified data fits into an existing classification. The classification unit 154 also determines whether the classification granularity of the classification result is correct. The classification unit 154 performs the classification process shown in Fig. 4 and Fig. 7.

[0100] The classification unit 154 classifies the plurality of data using related words related to the extracted keywords, which are keywords extracted from each of the plurality of data. The classification unit 154 classifies the plurality of data using related words, which are superordinate words of the extracted keywords, paraphrases that rephrase the extracted keywords in a different way, or associative words associated with the extracted keywords.

[0101] The classification unit 154 classifies multiple pieces of data based on the commonality of related words or extracted keywords. The classification unit 154 classifies multiple pieces of data by classifying data that share common related words or extracted keywords into the same group. The classification unit 154 classifies multiple pieces of data based on the related words identified by the keyword normalization unit 153. Note that the classification unit 154 may randomly classify multiple pieces of data as long as the criteria of data that share common related words or extracted keywords are met. Alternatively, the classification unit 154 may classify multiple pieces of data by starting with data that has an extracted keyword with a high score and grouping that data and data that share common related words or extracted keywords into the same group.

[0102] The classification unit 154 classifies a plurality of data into a plurality of groups. The classification unit 154 classifies a plurality of data into a plurality of groups using points of extracted keywords that indicate the degree of relevance of the extracted keywords to the data from which the extracted keywords were extracted. The classification unit 154 classifies a plurality of data into a plurality of groups using points of each extracted keyword stored in the classification target data storage unit 142. The classification unit 154 classifies a plurality of data into a plurality of groups using points of each group calculated using points of the extracted keywords corresponding to each of the plurality of groups. For example, the classification unit 154 may generate a plurality of division patterns for classifying the data into a plurality of groups, and use the division pattern with the highest average point value for each group from the plurality of division patterns. For example, the classification unit 154 may generate a set number of division patterns (e.g., 10 or 100) and use the division pattern with the highest average point value for each group from the plurality of division patterns. Note that the classification unit 154 may randomly generate division patterns if the data satisfy the criterion of having related words or extracted keywords in common. The classification unit 154 may also generate a division pattern by starting with data having an extracted keyword with a large point value and grouping that data and data having a related word or extracted keyword in common.

[0103] The classification unit 154 classifies the plurality of data into the plurality of groups if the points of each of the plurality of groups satisfy a predetermined criterion. The classification unit 154 classifies the plurality of data into the plurality of groups if the points of each of the plurality of groups are equal to or greater than a predetermined value. The classification unit 154 classifies the plurality of data into the plurality of groups if the points of each of the plurality of groups are equal to or greater than a point threshold stored in the classification condition storage unit 143.

[0104] The classification unit 154 classifies multiple data into groups in which the number of corresponding data items is equal to or less than a predetermined number. The classification unit 154 classifies the data so that the number of corresponding data items in each group is equal to or less than the maximum number of items stored in the classification condition storage unit 143. For example, if there is a group among the groups generated by the classification process in which the number of corresponding data items exceeds the maximum number of items, the classification unit 154 performs the classification process again. The classification unit 154 repeats the classification process until the number of corresponding data items in each of all groups classified by the classification process is equal to or less than the maximum number of items.

[0105] The classification unit 154 classifies a plurality of data into groups equal to or less than a classification number threshold. The classification unit 154 classifies the data so that the number of groups into which the data is classified is equal to or less than the maximum classification number stored in the classification condition storage unit 143. For example, if the number of groups resulting from the classification process exceeds the maximum classification number, the classification unit 154 performs the classification process again. The classification unit 154 repeats the classification process until the number of groups resulting from the classification becomes equal to or less than the maximum classification number.

[0106] The transmitting unit 155 transmits various types of information. The transmitting unit 155 transmits various types of information to an external information processing device. The transmitting unit 155 transmits various types of information to a terminal device used by a user (such as the user terminal 10 in FIG. 16). The transmitting unit 155 transmits information stored in the memory unit 14. The transmitting unit 155 transmits the classification results obtained by the classifying unit 154.

[0107] [1-3. Information Processing Procedure According to the Embodiment] Next, the procedure of information processing according to the embodiment will be described with reference to Fig. 12. Fig. 12 is a flowchart showing the procedure of information processing according to the embodiment of the present disclosure.

[0108] 12, the information processing device 100 extracts extracted keywords from each of a plurality of data to be classified (step S101). The information processing device 100 classifies the plurality of data based on the extracted keywords extracted from each of the plurality of data and related words related to the extracted keywords (step S102).

[0109] [1-4. Example of keyword extraction other than text] As described above, the data to be classified is not limited to text (character information), but may also be images, or time-series data such as video, audio, and waveforms. For example, the data to be classified may be sensing aggregate data (hereinafter also referred to as "sensing data") generated by sensor detection, such as video captured by an image sensor. Sensing may be any data generated by sensor detection, including waveform data and image data. FIG. 13 illustrates an example of keyword extraction other than text, using video as an example. FIG. 13 is a diagram showing an example of keyword extraction from data other than text. Note that explanations of points similar to those described above in FIGS. 1, 6, etc. will be omitted as appropriate.

[0110] The information processing device 100 extracts extracted keywords from sensing data (data DT11) which is video Y (step S51). The information processing device 100 extracts extracted keywords from video Y using video analysis technology. In FIG. 13, the information processing device 100 extracts extracted keywords such as "dog," "park," "Chihuahua," "Frisbee," and "run" from video Y, as shown in keyword information KY. In addition, the information processing device 100 identifies the part of speech of each extracted keyword and associates information on the identified part of speech with each extracted keyword.

[0111] The information processing device 100 calculates points for each extracted keyword of the video Y according to the degree of relevance with the video Y, and assigns the calculated points to each extracted keyword. For example, the information processing device 100 calculates points for each extracted keyword based on the time that an object corresponding to each extracted keyword appears in the video. The information processing device 100 calculates points for each extracted keyword such that the longer the time that an object appears in the video, the higher the point. The information processing device 100 may use the time that an object corresponding to each extracted keyword appears in the video as the point for each extracted keyword.

[0112] Furthermore, the information processing device 100 corrects (normalizes) the points assigned to the extracted keywords so that the total sum of the points is 1. In the example of Fig. 13, the information processing device 100 normalizes, for each extracted keyword of video Y, the points for "dog" to "0.6", the points for "park" to "0.2", the points for "chihuahua" to "0.1", the points for "frisbee" to "0.06", and the points for "run" to "0.04".

[0113] Then, the information processing device 100 normalizes the extracted keywords extracted from the sentence X (step S52). As shown in the normalized keyword information RKY, the information processing device 100 excludes "run" (run), which corresponds to the part of speech "verb" to be excluded, from the extracted keywords. Then, the information processing device 100 corrects (normalizes) the points assigned to each extracted keyword after excluding "run" so that the total sum of the points is 1. In the example of FIG. 13, the information processing device 100 normalizes the points of "dog" to "0.61", "park" to "0.21", "Chihuahua" to "0.11", and "Frisbee" to "0.07" for each extracted keyword of the video Y.

[0114] The information processing device 100 also identifies related words for each extracted keyword. Using a concept dictionary such as WordNet, related words are identified for each of "dog," "Chihuahua," and "Frisbee." In FIG. 13, as shown in the normalized keyword information RKY, the information processing device 100 identifies "pet" and "animal" as related words for "dog," "dog," "pet," and "animal" as related words for "Chihuahua," and "toy" as a related word for "Frisbee."

[0115] In this way, the information processing device 100 analyzes the video to extract information contained in the video. For example, the information processing device 100 extracts information contained in the video using a data analysis method such as object recognition within an image. For example, the information processing device 100 may divide the video into frame images, apply object recognition to each, and calculate points based on the ratio of the number of frames in which an object (target) is recognized to the total number of frames. Furthermore, for example, the information processing device 100 may use time-series information as keywords for video image recognition. Among the keywords extracted in FIG. 13, "run" corresponds to time-series keyword information. Furthermore, the information processing device 100 may extract audio information from the video and perform audio analysis. In this case, the information processing device 100 may apply voice recognition and extract keywords from meaningful character strings in the same way as text analysis. Furthermore, the information processing device 100 may classify audio data, such as animal cries or outdoors. Furthermore, the information processing device 100 may perform these voice analyses and determine extracted keywords in combination with keywords obtained from the video. Note that the above is just an example, and the information processing device 100 may extract extracted keywords from data such as video using other methods.

[0116] [1-5. Classification update example] For example, if newly added data does not fall into an existing classification (group), the classification of the existing data may be reconsidered. In such a case, the information processing device 100 may update the classification by performing the classification process again. An example of updating the classification will be described with reference to FIGS. 14 and 15. FIG. 14 is a diagram showing an example of updating the classification. FIG. 15 is a diagram showing an example of the result of updating the classification. Note that, for the sake of simplicity, FIGS. 14 and 15 show an example using only extracted keywords, but the information processing device 100 may also perform the classification process using information on related words of each extracted keyword.

[0117] In Figure 14, as shown in the classification result RS21, sentences A, B, and G are classified into group G21 corresponding to the classification word "sports," and sentences C, D, and H are classified into group G12 corresponding to the classification word "pets."

[0118] Here, a new sentence I is added. The information processing device 100 extracts extracted keywords from the sentence I as shown in the keyword information KI, and calculates the points for each extracted keyword. For each extracted keyword in the sentence I, the information processing device 100 normalizes the points for "recipe" to "0.4", "time saving" to "0.3", "prepared meal" to "0.2", "microwave" to "0.06", and "easy" to "0.04".

[0119] Here, since the extracted keywords of sentence I do not include "sports" or "pets," the information processing device 100 determines whether to exclude sentence I from the classification or to update the overall classification by including sentence I. As shown in Fig. 14, sentences G, H, and I share the abstract keyword "recipe." Therefore, the information processing device 100 generates, as a classification candidate, a classification result RS22 in which sentences G, H, and I are classified into one group.

[0120] In the case of the classification result RS22 shown in Figure 15, sentences G, H, and I are classified into group G26 corresponding to the classification word "recipes." Sentences A and B are classified into group G24 corresponding to the classification word "sports," and sentences C, D, and E are classified into group G25 corresponding to the classification word "pets."

[0121] In the case of classification result RS22, the points for group G26 corresponding to the classification word "recipes" are "0.3", the points for group G24 corresponding to the classification word "sports" are "0.25", and the points for group G25 corresponding to the classification word "pets" are "0.3".

[0122] The information processing device 100 determines whether to update the classification using the average value of the points for each group in each classification result. The information processing device 100 determines whether to update the classification based on the average value of the points for each group in the classification at the time of determination and the average value of the points for each group in the generated classification candidates. In this way, the information processing device 100 determines whether to update the classification based on the average value of the points for each group when new sentence I is not added to the classification and the average value of the points for each group when new sentence I is added to the classification target.

[0123] 14 and 15, the information processing device 100 determines whether to update the classification based on the comparison result between the average point value of each group in classification result RS21, which is the classification at the time of determination, and the average point value of each group in classification result RS22, which is a classification candidate. Specifically, the information processing device 100 determines to update the classification if the average point value of each group in classification result RS22 is equal to or greater than the average point value of each group in classification result RS21.

[0124] Here, as shown in the classification result RS21, if the new sentence I is not added to the classification, the points of the group G21 corresponding to the classification word "sports" will be "0.2", and the points of the group G22 corresponding to the classification word "pets" will be "0.3". Therefore, the information processing device 100 calculates the average value of the points of the groups G21 and G22 in the classification result RS21 (hereinafter referred to as "the points of the classification result RS21") as "0.25 (= (0.2 + 0.3) / 2)".

[0125] Furthermore, as shown in the classification result RS22, when a new sentence I is added to the classification, the points of the group G24 corresponding to the classification word "sports" become "0.25", the points of the group G25 corresponding to the classification word "pets" become "0.3", and the points of the group G26 corresponding to the classification word "recipes" become "0.3". Therefore, the information processing device 100 calculates the average value of the points of the groups G24, G25, and G26 in the classification result RS22 (hereinafter referred to as "points of the classification result RS22") as "0.283 (= (0.25 + 0.3 + 0.3) / 3)".

[0126] The information processing device 100 compares the score "0.25" of the classification result RS21 with the score "0.283" of the classification result RS22. Then, since the score of the classification result RS22 is equal to or greater than the score of the classification result RS21, the information processing device 100 determines that the classification of the classification result RS22 is more in line with the content of the data, and decides to update the classification. In this case, the information processing device 100 updates the classification from the classification result RS21 to the classification result RS22 (step S61).

[0127] The above is just an example, and the information processing device 100 may determine whether to update the classification using various information other than the average value of the points of each group in each classification.

[0128] [1-6.Classification conditions] The information processing device 100 may perform classification based on various classification conditions, not limited to the above example. Each piece of data may belong to multiple groups. In the example of FIG. 15, sentence G may belong to two groups: group G24 corresponding to the classification word "sports" and group G26 corresponding to the classification word "recipes." Furthermore, sentence H may belong to two groups: group G25 corresponding to the classification word "pets" and group G26 corresponding to the classification word "recipes." In this case, the information processing device 100 classifies each piece of data by allowing each piece of data to belong to multiple groups. Note that an upper limit (for example, 3 or 5) may be set for the number of groups to which each piece of data belongs.

[0129] Furthermore, the information processing device 100 may set a maximum number of categories instead of the maximum number of items, or may use machine learning to further optimize the number of categories (number of groups) and the number of items within a category (number of data included in a group).

[0130] [2. Other embodiments] The processing according to each of the above-described embodiments may be implemented in various different forms (modifications) other than the above-described embodiments. For example, the system configuration is not limited to the above-described examples, and may be in various forms. This point will be described below. Note that, in the following, explanations of points similar to those of the information processing device 100 according to the embodiment will be omitted as appropriate.

[0131] [2-1. Modifications] For example, instead of performing the classification process completely automatically, one or more patterns (also called "classification candidates") may be presented to the user as candidate classification methods. The pattern selected by the user may then be used as the classification result. This allows the user to select a classification that is more in line with their own intuition, and the execution results may be incorporated again into the machine learning data as the user's preferences. Furthermore, the user may be allowed to make a selection regarding classification when a certain threshold is exceeded. Furthermore, the user's selection may be determined to be the user's preference and reflected in subsequent automatic classification results.

[0132] An example configuration of the information processing system 1 in the case where the user is allowed to select a category as described above will be described with reference to Fig. 16. Fig. 16 is a diagram illustrating an example configuration of an information processing system according to a modified example of the present disclosure.

[0133] 16, the information processing system 1 includes a user terminal 10 and an information processing device 100. The user terminal 10 and the information processing device 100 are connected to each other via a communication network N so as to be able to communicate with each other via a wired or wireless connection. Note that the information processing system 1 shown in FIG. 16 may include a plurality of user terminals 10 and a plurality of information processing devices 100. For example, the information processing system 1 may include a plurality of user terminals 10, such as a user terminal 10 used by a user U1 and a user terminal 10 used by a user U2.

[0134] The information processing device 100 generates multiple classification results for multiple data to be classified. Then, the information processing device 100 transmits the multiple classification results as multiple classification candidates to the user terminal 10. The information processing device 100 receives information indicating the classification candidates selected by the user from the user terminal 10. The information processing device 100 associates the received information indicating the classification candidates selected by the user with the user and stores it in the storage unit 120 as the user's selection history. Then, the information processing device 100 generates a classification result for the user using the user's selection history.

[0135] The user terminal 10 is an information processing device used by a user. The user terminal 10 is realized by, for example, a smartphone, a tablet terminal, a smart speaker, a mobile phone, a PDA (Personal Digital Assistant), a notebook PC (Personal Computer), a desktop PC, a smartphone, etc. The user terminal 10 may be any terminal device that can display information provided by the information processing device 100 and transmit information indicating a classification selected by the user to the information processing device 100.

[0136] Furthermore, the user terminal 10 accepts operations by the user. The user terminal 10 displays the information received from the information processing device 100 on the screen. The user terminal 10 displays the classification results received from the information processing device 100.

[0137] The user terminal 10 displays the multiple classification candidates provided by the information processing device 100. The user terminal 10 also accepts a user's selection from the multiple classification candidates and transmits information indicating the classification candidate selected by the user to the information processing device 100.

[0138] [2-1-1. Classification by user] Here, classification according to users will be described with reference to FIG. 17. FIG. 17 is a diagram showing an example of classification according to users. FIG. 17 takes two users, user U1 and user U2, as examples, and shows an example in which as users select classifications, the user preferences are reflected and the subsequent classification differs for each user. FIG. 17 shows 11 pieces of data, sentences A to L. Furthermore, keyword information KA to keyword information KL correspond to sentences A to L, respectively. Note that related words are not shown in FIG. 17. For example, the extracted keyword "cake" of sentence H shown in keyword information KH is associated with the related word "recipe."

[0139] Classification results RS31 and RS41 in FIG. 17 show classification results for eight data items, sentences A, B, C, D, E, G, H, and I.

[0140] The classification result RS31 indicates the classification result selected by the user U1. The classification result RS31 is a classification result that classifies the sentence G with the keywords "muscle training" and "recipe" and the sentence H with the keywords "cake" and "pet" into the group G33 corresponding to the classification word "recipe." The information processing device 100 reflects the selection of the user U1 as the preference of the user U1 in the classification process when providing the information to the user U1.

[0141] The classification result RS41 indicates the classification result selected by the user U2. The classification result RS41 classifies the sentence G containing the keywords "muscle training" and "recipe" into the group G41 corresponding to the classification word "sports," and classifies the sentence H containing the keywords "cake" and "pet" into the group G42 corresponding to the classification word "recipe."

[0142] The information processing device 100 reflects the selection of the user U2 as the preference of the user U2 in the classification process when providing the classification result to the user U2. The information processing device 100 corrects the classification so as to reflect the preference of the user U2.

[0143] For example, in the classification process when providing the classification results to user U2, the information processing device 100 may prioritize classification into a group related to sports or pets, even if the data is recipe-related. Furthermore, for example, the information processing device 100 may assume that user U2 has little interest in recipes, and correct data containing recipe-related keywords so that the data is classified into a category selected by user U2, rather than recipes. In this case, the information processing device 100 may assign lower points to recipe-related keywords.

[0144] As a result, when new data is added, the information processing device 100 may generate different classification results for the user U1 and the user U2.

[0145] When three new data items, Sentence J, Sentence K, and Sentence L, are added, the information processing device 100 classifies the user U1 as in classification result RS32 (step S71). When three new data items, Sentence J, Sentence K, and Sentence L, are added, the information processing device 100 classifies the user U2 as in classification result RS42 (step S81).

[0146] 17 shows a case where a sentence J containing the keywords "hot pot" and "sumo" is classified differently. For user U1 who has not made any corrections, the information processing device 100 classifies the sentence J into group G36 corresponding to the classification word "recipe."

[0147] On the other hand, for the corrected user U2, the information processing device 100 does not classify the text J as recipe-related. In the example of Fig. 17, for the user U2, the information processing device 100 classifies the text J into group G44 corresponding to the classification word "sports".

[0148] Similarly, a sentence K having the keywords "diet recipe" and "dog" is classified into group G36 corresponding to the classification word "recipe" for user U1, but into group G45 corresponding to the classification word "pet" for user U2.

[0149] [2-2. Other configuration examples] Furthermore, the processes according to the above-described embodiments and modifications may be implemented in various different forms (modifications) other than the above-described embodiments and modifications. For example, the information processing device 100 may be integrated with the user terminal 10. In this way, the user terminal 10 used by each user may have the functions of the information processing device 100. Note that the above is just an example, and the information processing system may be realized in various configurations.

[0150] [2-3.Other] Furthermore, among the processes described in the above embodiments, all or part of the processes described as being performed automatically can be performed manually, or all or part of the processes described as being performed manually can be performed automatically using known methods. Furthermore, the information, including the processing procedures, specific names, various data, and parameters shown in the above documents and drawings, can be changed as desired unless otherwise specified. For example, the various information shown in each drawing is not limited to the information shown in the drawings.

[0151] Furthermore, the components of each device shown in the figure are conceptual functional components and do not necessarily have to be physically configured as shown in the figure. In other words, the specific form of distribution and integration of each device is not limited to that shown in the figure, and all or part of them can be functionally or physically distributed and integrated in any unit depending on various loads, usage conditions, etc.

[0152] Furthermore, the above-described embodiments and modifications can be combined as appropriate within the scope of not causing any contradiction in the processing content.

[0153] Furthermore, the effects described in this specification are merely examples and are not limiting, and other effects may also be present.

[0154] [3. Effects of this disclosure] As described above, the information processing device according to the present disclosure (information processing device 100 in the embodiment) includes a keyword extraction unit (keyword extraction unit 152 in the embodiment) and a classification unit (classification unit 154 in the embodiment). The keyword extraction unit extracts keywords from each of a plurality of data to be classified. The classification unit classifies the plurality of data using related words related to the extracted keywords, which are keywords extracted from each of the plurality of data.

[0155] As a result, the information processing device according to the present disclosure can classify data by taking into account information on related words related to the extracted keywords extracted from multiple data that are the classification format, thereby enabling appropriate classification of data.

[0156] The classification unit classifies the plurality of data using hypernyms of the extracted keywords, paraphrases that rephrase the extracted keywords, or related words that are associated words associated with the extracted keywords. This allows the information processing device to classify the data using hypernyms, paraphrases, or associated words of the extracted keywords, taking into account the conceptual commonality of the extracted keywords, thereby enabling appropriate classification of the data.

[0157] The classification unit classifies the plurality of data based on the commonality of the related words or extracted keywords. This allows the information processing device to classify the data taking into account the commonality of the related words or extracted keywords, thereby enabling the data to be appropriately classified.

[0158] The information processing device according to the present disclosure also includes a keyword normalization unit (keyword normalization unit 153 in the embodiment). The keyword normalization unit identifies related words of the extracted keyword and associates them with the data from which the extracted keyword was extracted. The classification unit classifies multiple data sets based on the related words identified by the keyword normalization unit. This allows the information processing device to identify related words of the extracted keyword and appropriately classify the data using the related words associated with the data from which the extracted keyword was extracted.

[0159] The keyword normalization unit also identifies related words for the extracted keywords using a concept dictionary that indicates conceptual relationships between languages, allowing the information processing device to appropriately identify related words for the extracted keywords based on the conceptual relationships between languages.

[0160] Furthermore, the keyword normalization unit excludes keywords that satisfy the exclusion condition from the extracted keywords. This allows the information processing device to classify data by excluding extracted keywords that are unnecessary for classification processing, thereby enabling data to be appropriately classified.

[0161] The keyword normalization unit also excludes keywords that correspond to excluded parts of speech from the extracted keywords. This allows the information processing device to classify data groups while excluding extracted keywords that correspond to parts of speech that are not required for classification processing, thereby enabling data to be appropriately classified.

[0162] The classification unit classifies the plurality of data into a plurality of groups. This allows the information processing device to classify the data groups taking into account the commonality of the related words or extracted keywords, thereby enabling the data to be appropriately classified.

[0163] The classification unit classifies the plurality of data into a plurality of groups using points of the extracted keywords that indicate the degree of relevance of the extracted keywords to the data from which the extracted keywords were extracted. This allows the information processing device to classify the data groups taking into account the degree of relevance of the extracted keywords to the data, thereby enabling appropriate classification of the data.

[0164] The classification unit classifies the plurality of data using the points of each of the plurality of groups calculated using the points of the extracted keywords corresponding to each of the plurality of groups. This allows the information processing device to classify the data group taking into account the points of each group into which the data is classified, thereby enabling the data to be classified appropriately.

[0165] Furthermore, the classification unit classifies the plurality of data into the plurality of groups when the points of each of the plurality of groups satisfy a predetermined criterion. As a result, the information processing device can appropriately classify the data by classifying the data into the group when the points of each group into which the data is classified satisfy the criterion.

[0166] Furthermore, the classification unit classifies the plurality of data into the plurality of groups when the points of each of the plurality of groups are equal to or greater than a predetermined value. As a result, the information processing device can appropriately classify the data by classifying the data into the group when the points of each group into which the data is classified are equal to or greater than a predetermined value.

[0167] The classification unit classifies the plurality of data into groups in which the number of corresponding data is equal to or less than a predetermined number. As a result, the information processing device can prevent a large number of data from being included in one group by classifying the data so that the number of data belonging to each group does not exceed a predetermined value, thereby enabling the data to be classified appropriately.

[0168] Furthermore, the classification unit classifies a plurality of data into groups with a number of classifications equal to or smaller than the classification number threshold. This allows the information processing device to prevent data from being classified into a large number of groups, and to appropriately classify data.

[0169] Furthermore, the plurality of data includes text data, and thus the information processing device can appropriately classify the data by using the plurality of data including the text data as the classification target.

[0170] Furthermore, the plurality of data includes image data, and thus the information processing device can appropriately classify the plurality of data including image data as the classification target.

[0171] Furthermore, the plurality of data includes time-series data, which allows the information processing device to appropriately classify the plurality of data including time-series data as the classification target.

[0172] Furthermore, the time-series data is moving image data, audio data, or waveform data, and the information processing device can appropriately classify the data by classifying a plurality of data including moving image data, audio data, or waveform data.

[0173] [4. Hardware Configuration] An information device such as the information processing device 100 according to each of the above-described embodiments is realized by, for example, a computer 1000 configured as shown in FIG. 18. FIG. 18 is a hardware configuration diagram showing an example of the computer 1000 that realizes the functions of an information processing device such as the information processing device 100. The information processing device 100 according to the embodiment will be described below as an example. The computer 1000 has a CPU 1100, a RAM 1200, a ROM (Read Only Memory) 1300, a HDD (Hard Disk Drive) 1400, a communication interface 1500, and an input / output interface 1600. The components of the computer 1000 are connected by a bus 1050.

[0174] The CPU 1100 operates and controls each unit based on programs stored in the ROM 1300 or the HDD 1400. For example, the CPU 1100 loads the programs stored in the ROM 1300 or the HDD 1400 into the RAM 1200 and executes processing corresponding to the various programs.

[0175] The ROM 1300 stores boot programs such as a Basic Input Output System (BIOS) executed by the CPU 1100 when the computer 1000 is started, and programs that depend on the hardware of the computer 1000 .

[0176] HDD 1400 is a computer-readable recording medium that non-temporarily records programs executed by CPU 1100 and data used by such programs. Specifically, HDD 1400 is a recording medium that records an information processing program according to the present disclosure, which is an example of program data 1450.

[0177] The communication interface 1500 is an interface for connecting the computer 1000 to an external network 1550 (e.g., the Internet). For example, the CPU 1100 receives data from other devices and transmits data generated by the CPU 1100 to other devices via the communication interface 1500.

[0178] The input / output interface 1600 is an interface for connecting the input / output device 1650 and the computer 1000. For example, the CPU 1100 receives data from an input device such as a keyboard or a mouse via the input / output interface 1600. The CPU 1100 also transmits data to an output device such as a display, a speaker, or a printer via the input / output interface 1600. The input / output interface 1600 may also function as a media interface for reading programs and the like recorded on a predetermined recording medium. Examples of media include optical recording media such as a DVD (Digital Versatile Disc) or a PD (Phase Change Rewritable Disk), magneto-optical recording media such as an MO (Magneto-Optical disk), tape media, magnetic recording media, and semiconductor memories.

[0179] For example, when the computer 1000 functions as the information processing device 100 according to the embodiment, the CPU 1100 of the computer 1000 executes an information processing program loaded onto the RAM 1200 to realize functions of the control unit 15 and the like. The information processing program according to the present disclosure and data in the storage unit 14 are stored in the HDD 1400. The CPU 1100 reads and executes program data 1450 from the HDD 1400, but as another example, the CPU 1100 may obtain these programs from another device via an external network 1550.

[0180] The present technology can also be configured as follows. (1) a keyword extraction unit that extracts keywords from each of a plurality of data to be classified; a classification unit that classifies the plurality of data using related words related to extracted keywords, which are keywords extracted from each of the plurality of data; An information processing device comprising: (2) The classification unit classifying the plurality of data using related words, which are hypernyms of the extracted keywords, paraphrases obtained by rephrasing the extracted keywords in other words, or associated words associated with the extracted keywords; An information processing device according to (1). (3) The classification unit classifying the plurality of data based on commonalities of the related words or the extracted keywords; An information processing device according to (1) or (2). (4) The classification unit classifying the plurality of data by classifying data having the common related words or extracted keywords into the same group; (3) An information processing device according to the present invention. (5) a keyword normalization unit that identifies the related words of the extracted keywords and associates the related words with the data from which the extracted keywords were extracted; Equipped with The classification unit classifying the plurality of data based on the related words identified by the keyword normalization unit; The information processing device according to any one of (1) to (4). (6) The keyword normalization unit Identifying the related words of the extracted keywords using a concept dictionary that indicates conceptual relationships between languages; (5) An information processing device according to (5). (7) The keyword normalization unit Excluding keywords that satisfy an exclusion condition from the extracted keywords; An information processing device according to (5) or (6). (8) The keyword normalization unit Excluding keywords that fall under the excluded parts of speech from the extracted keywords; (7) An information processing device according to (7). (9) The classification unit classifying the plurality of data into a plurality of groups; The information processing device according to any one of (1) to (8). (10) The classification unit classifying the plurality of data into the plurality of groups using points of the extracted keywords that indicate the degree of relevance of the extracted keywords to the data from which the extracted keywords were extracted; (9) An information processing device according to (9). (11) The classification unit classifying the plurality of data using points for each of the plurality of groups calculated using points for the extracted keywords corresponding to each of the plurality of groups; (10) An information processing device according to (10). (12) The classification unit classifying the plurality of data into the plurality of groups if each point in the plurality of groups meets a predetermined criterion; (11) An information processing device according to (11). (13) The classification unit classifying the plurality of data into the plurality of groups when the points of each of the plurality of groups are equal to or greater than a predetermined value; (12) An information processing device according to (12). (14) The classification unit classifying the plurality of data into groups in which the number of corresponding data is equal to or less than a predetermined number; The information processing device according to any one of (1) to (13). (15) The classification unit classifying the plurality of data into groups equal to or smaller than a classification number threshold; The information processing device according to any one of (1) to (14). (16) The plurality of data includes: Including text data, The information processing device according to any one of (1) to (15). (17) The plurality of data includes: including image data, The information processing device according to any one of (1) to (16). (18) The plurality of data includes: including time series data, The information processing device according to any one of (1) to (17). (19) The time series data is video data, audio data, or waveform data; (18) An information processing device according to (18). (20) Extracting keywords from each of the multiple data to be classified, classifying the plurality of data using related words related to extracted keywords, which are keywords extracted from each of the plurality of data; An information processing method that performs processing. [Explanation of symbols]

[0181] 100 Information processing device 11 Communications Department 12 Input section 13 Display unit (display) 14 Storage section 141 Concept Dictionary Memory Unit 142 Classification target data storage unit 143 Classification condition storage unit 15 Control Unit 151 Acquisition Department 152 Keyword Extraction Unit 153 Keyword normalization section 154 Classification Department 155 Transmitter

Claims

1. a keyword extraction unit that extracts keywords from each of a plurality of data to be classified; a classification unit that classifies the plurality of data using related words related to extracted keywords, which are keywords extracted from each of the plurality of data; a keyword normalization unit that identifies the related words of the extracted keywords and associates the related words with the data from which the extracted keywords were extracted; Equipped with The classification unit An information processing device that classifies the plurality of data based on the related words identified by the keyword normalization unit.

2. The classification unit classifying the plurality of data using related words, which are hypernyms of the extracted keywords, paraphrases obtained by rephrasing the extracted keywords in other words, or associated words associated with the extracted keywords; The information processing device according to claim 1 .

3. The classification unit classifying the plurality of data based on commonalities of the related words or the extracted keywords; The information processing device according to claim 1 .

4. The classification unit classifying the plurality of data by classifying data having the common related words or extracted keywords into the same group; The information processing device according to claim 3 .

5. The keyword normalization unit Identifying the related words of the extracted keywords using a concept dictionary that indicates conceptual relationships between languages; The information processing device according to claim 1 .

6. The keyword normalization unit Excluding keywords that satisfy an exclusion condition from the extracted keywords; The information processing device according to claim 1 .

7. The keyword normalization unit Excluding keywords that fall under the excluded parts of speech from the extracted keywords; The information processing device according to claim 6 .

8. A keyword extraction unit that extracts keywords from each of a plurality of data to be classified; a classification unit that classifies the plurality of data using related words related to extracted keywords, which are keywords extracted from each of the plurality of data; Equipped with The classification unit An information processing device that classifies the plurality of data into a plurality of groups using points of the extracted keywords that indicate the degree of relevance of the extracted keywords to the data from which the extracted keywords were extracted.

9. The classification unit classifying the plurality of data using points for each of the plurality of groups calculated using points for the extracted keywords corresponding to each of the plurality of groups; The information processing device according to claim 8 .

10. The classification unit classifying the plurality of data into the plurality of groups if each point in the plurality of groups meets a predetermined criterion; The information processing device according to claim 9 .

11. The classification unit classifying the plurality of data into the plurality of groups when the points of each of the plurality of groups are equal to or greater than a predetermined value; The information processing device according to claim 10.

12. The classification unit classifying the plurality of data into groups in which the number of corresponding data is equal to or less than a predetermined number; The information processing device according to claim 1 .

13. The classification unit classifying the plurality of data into groups equal to or smaller than a classification number threshold; The information processing device according to claim 1 .

14. The plurality of data includes: Including text data, The information processing device according to claim 1 .

15. The plurality of data includes: including image data, The information processing device according to claim 1 .

16. The plurality of data includes: including time series data, The information processing device according to claim 1 .

17. The time series data is video data, audio data, or waveform data; The information processing device according to claim 16.

18. A computer-implemented information processing method, comprising: Extracting keywords from each of the multiple data to be classified, classifying the plurality of data using related words related to extracted keywords, which are keywords extracted from each of the plurality of data; Identifying the related words of the extracted keywords and associating them with the data from which the extracted keywords were extracted; Execute the process, An information processing method for classifying the plurality of data based on the related words identified by the computer.

19. A computer-implemented information processing method, comprising: Extracting keywords from each of the multiple data to be classified, classifying the plurality of data using related words related to extracted keywords, which are keywords extracted from each of the plurality of data; Execute the process, An information processing method in which the computer classifies the plurality of data into a plurality of groups using points of the extracted keywords that indicate the degree of relevance of the extracted keywords to the data from which they were extracted.

Images