Information processing device and information processing method

Abstract

An information processing device (100) comprises an extracting unit (120), a collecting unit (130), and a determining unit (140). The extracting unit extracts content on the basis of an extraction condition relating to target content for each of a plurality of pieces of content information relating to the target content, and generates a set of elements relating to the content as extraction information. The collecting unit classifies a plurality of elements included in two different pieces of extraction information among the plurality of pieces of generated extraction information into a plurality of categories, and generates, for the plurality of categories, a union set composed of a common set and two difference sets as union set information, the common set being a set of categories containing elements included in each of the two pieces of extraction information, and the two difference sets being sets of categories including elements included in either of the two pieces of extraction information. The determining unit determines the order of the plurality of pieces of content information on the basis of the plurality of pieces of generated union set information corresponding to all combinations of the plurality of pieces of extracted information, and generates order information indicating the order of the plurality of pieces of content information.

Description

Information processing device and information processing method

[0001] The present invention relates to an information processing device and an information processing method.

[0002] For example, in employee training, there are cases where the training materials are not specified by the instructor, and learners themselves must select multiple materials and create a learning plan based on those selected materials. In such cases, in order to conduct effective learning, it is important for learners to use multiple materials in an order that is effective for learning.

[0003] Conventional technologies propose methods for recommending multiple books based on the user's intended use and information about the courses they are taking. In these conventional methods, the list of recommended books is presented in order of how closely it matches the student's preferences.

[0004] Ryūhei Miyazato, Tetsuro Nishino, and Kei Harada, "A Proposal for a Course-Related Book Recommendation Method Considering the Learning Status of University Library Users," Information Processing Society of Japan Research Report, June 21, 2024, Vol. 2024-MPS-148 No. 46.

[0005] As described above, the conventional method presents multiple books included in the book list (which are included in the broader definition of "multiple content") in an order that is closest to the learner's preference. Therefore, the conventional method may not necessarily present a content usage order that allows for efficient knowledge acquisition.

[0006] The present invention aims to provide a technology that presents a content usage order that enables efficient knowledge acquisition.

[0007] An information processing device according to one aspect of the present invention comprises an extraction unit, an aggregation unit, and a determination unit. The extraction unit extracts content from each of a plurality of content information relating to the target content based on extraction conditions relating to the target content, and generates a set of elements relating to the content as extracted information. For every combination of two different extracted pieces of information corresponding to each of the plurality of content information, the aggregation unit classifies the elements contained in the two extracted pieces of information into a plurality of categories based on classification conditions relating to the target content, and generates union information for the plurality of categories, which consists of a common set, which is a set of categories containing elements contained in each of the two extracted pieces of information, and a difference set, which is a set of categories containing elements contained in either of the two extracted pieces of information. The determination unit determines the ordering of the plurality of content information based on the generated union information corresponding to each of the plurality of extracted pieces of information, and generates ordering information indicating the order of the plurality of content information.

[0008] According to the present invention, a technology is provided that presents a sequence for using content that enables efficient knowledge acquisition.

[0009] Figure 1 is a block diagram showing an example of an overview of the sequence presentation process by the information processing device according to the embodiment. Figure 2 is a block diagram showing an example of the hardware configuration of the information processing device in Figure 1. Figure 3 is a block diagram showing an example of the functional configuration of the information processing device in Figure 2. Figure 4 is a flowchart showing an example of the operation of the sequence presentation process in the information processing device according to the embodiment. Figure 5 is an explanatory diagram of the word extraction process and abstraction process in the first specific example of the embodiment. Figure 6 is an explanatory diagram of the set processing in the first specific example of the embodiment. Figure 7 is an explanatory diagram of the sequence determination process in the first specific example of the embodiment. Figure 8 is an explanatory diagram of the word extraction process, image extraction process, and vectorization process in the second specific example of the embodiment. Figure 9 is an explanatory diagram of the set processing in the second specific example of the embodiment. Figure 10 is an explanatory diagram of the sequence determination process in the second specific example of the embodiment. Figure 11 is an explanatory diagram of the content in the third specific example of the embodiment. Figure 12 is an explanatory diagram of the classification process and abstraction process in the third specific example of the embodiment. Figure 13 is an explanatory diagram of the set processing in the third specific example of the embodiment. Figure 14 is an explanatory diagram of the sequence determination process in the third specific example of the embodiment. Figure 15 is an explanatory diagram of the word extraction process and vectorization process in the fourth specific example of the embodiment. Figure 16 is an explanatory diagram of the set processing in the fourth specific example of the embodiment. Figure 17 is an explanatory diagram of the order determination process in the fourth specific example of the embodiment.

[0010] Embodiments of the present invention will be described below with reference to the drawings.

[0011] <Embodiment> [Configuration] Figure 1 is a block diagram showing an example of an overview of the sequence presentation process by an information processing device according to the embodiment. The information processing device 100 in Figure 1 can generate sequence information that presents multiple content pieces in a predetermined order when it receives multiple content pieces related to the target content as input.

[0012] The target content includes, for example, educational materials, manuals, books, and practice problems that users use to acquire knowledge. Specifically, the target content is expected to include, for example, "programming materials," "operation manuals," "visual programming practice problems," and "books of any genre."

[0013] Content information refers to information that includes the content of the target content. Specifically, the content of the target content includes text data such as strings of characters, and visual data such as images and videos. In the following explanation, multiple pieces of content information refer to different pieces of content information.

[0014] Figure 2 is a block diagram showing an example of the hardware configuration of the information processing device shown in Figure 1. The information processing device 100 in Figure 2 includes a control circuit 101, storage 102, communication module 103, interface 104, and drive 105.

[0015] The control circuit 101 is a circuit that controls all components of the information processing device 100 as a whole. The control circuit 101 includes a CPU (Central Processing Unit), RAM (Random Access Memory), ROM (Read Only Memory), etc. The ROM of the control circuit 101 stores programs used for various processes in the information processing device 100. The CPU of the control circuit 101 controls the entire information processing device 100 according to the programs stored in the ROM of the control circuit 101. The RAM of the control circuit 101 is used as a work area for the CPU of the control circuit 101.

[0016] The storage 102 is composed of, for example, an HDD (Hard Disk Drive), an SSD (Solid State Drive), or flash memory. Information used for various processes in the information processing device 100 is stored in the storage 102.

[0017] The communication module 103 is a circuit used for sending and receiving data with other devices (external devices).

[0018] Interface 104 is an interface primarily used for communication between the user and the control circuit 101. Interface 104 includes, for example, input devices and output devices. Input devices include, for example, a microphone, camera, touch panel, and operation buttons. Output devices include, for example, a speaker and a display.

[0019] Drive 105 is a device for reading software stored on the storage medium 106. Drive 105 includes, for example, a CD (Compact Disk) drive or a DVD (Digital Versatile Disk) drive.

[0020] The storage medium 106 is a medium for storing software by electrical, magnetic, optical, mechanical, or chemical means. The storage medium 106 may also store programs for executing various processes in the information processing device 100.

[0021] The storage medium 106 may also be a USB (Universal Serial Bus) memory. If the storage medium 106 is a USB memory, it is connected to, for example, the communication module 103. In this case, the information processing device 100 does not need to have a drive 105.

[0022] Figure 3 is a block diagram showing an example of the functional configuration of the information processing device shown in Figure 2. The information processing device 100 in Figure 3 includes an acquisition unit 110, an extraction unit 120, an aggregation unit 130, and a determination unit 140.

[0023] To realize the functional configuration shown in Figure 3, the CPU of the control circuit 101 in Figure 2 loads the program stored in the ROM or storage medium 106 of the control circuit 101 into the RAM of the control circuit 101. The CPU of the control circuit 101 then interprets and executes the program loaded into the RAM of the control circuit 101. As a result, the information processing device 100 functions as a computer comprising an acquisition unit 110, an extraction unit 120, an aggregation unit 130, and a determination unit 140.

[0024] The acquisition unit 110 acquires multiple content information items related to the target content through user operation. The acquisition unit 110 may also acquire extraction conditions and classification conditions related to the target content. The extraction conditions are used in the extraction unit 120 and will be described later. The classification conditions are used in the aggregation unit and will be described later. The acquisition unit 110 may acquire this information, conditions, etc., from, for example, storage 102 or an external device.

[0025] The extraction unit 120 extracts the content contained in each of the multiple content information acquired by the acquisition unit 110 based on the extraction conditions. For each of the multiple content information, the extraction unit 120 generates a set of elements related to the content as extracted information. That is, the extraction unit 120 generates multiple pieces of extracted information corresponding to each of the multiple content information.

[0026] The extraction criteria may be predetermined according to the target content. Specifically, the extraction criteria include text data, visual data, etc., contained in the content information related to the target content. Therefore, the extracted content will include text data, visual data (image data, video data, etc.), etc., contained in the content information.

[0027] The elements related to the content include the extracted text data itself, data (information) obtained by applying various processes (filtering, vectorization, abstraction, etc.) to the extracted text data, and data (information) obtained by applying various processes (classification, vectorization, abstraction, etc.) to the extracted visual data. The extraction unit 120 may generate the above elements by applying various processes to the text data, visual data, etc. contained in each of the multiple content information items.

[0028] The aggregation unit 130 classifies a plurality of elements included in two pieces of extraction information into a plurality of categories based on a classification condition, for each of all combinations of two different pieces of extraction information among the plurality of pieces of extraction information generated by the extraction unit 120. For the plurality of classified categories, the aggregation unit 130 generates, as union set information, a union set consisting of a common set that is a set of categories having elements included in each of the two pieces of extraction information, and two difference sets that are sets of categories having elements included in either one of the two pieces of extraction information. That is, the aggregation unit 130 generates a plurality of pieces of union set information each corresponding to all combinations of two different pieces of extraction information among the plurality of pieces of extraction information.

[0029] In the following, the two pieces of extraction information are respectively referred to as extraction information A １ and extraction information A ２ , the union set is represented by "A １ ∪A ２ ", the common set is represented by "A １ ∩A ２ ", the difference set obtained by subtracting A １ from A ２ is represented by "A １ \A ２ ", and the difference set obtained by subtracting A ２ from A １ is represented by "A ２ \A １ ". Further, in the union set A １ ∪A ２ , the difference set A １ \A ２ may be referred to as "one difference set", and the difference set ２ \A １ may be referred to as "the other difference set".

[0030] The classification condition may be determined in advance according to the target content. For example, terms related to the target content, the distance between elements after individual data is vectorized, the type of material constituting the target content, and the like may be used for the classification condition. Further, the number of categories classified based on the classification condition may be determined in advance.

[0031] The set unit 130 may generate union information with weights or priorities assigned to the categories. Alternatively, the set unit 130 may create union information before identifying the intersection and the difference between the two sets, and then identify the intersection and the difference between the two sets.

[0032] The determination unit 140 determines the ordering of multiple content information based on the multiple union information generated by the set unit 130, and generates order information indicating the order of the multiple content information. Specifically, the determination unit 140 may identify the number of categories or elements included in one of the difference sets for each of the multiple union information, and determine that the content information corresponding to the difference set with the smallest number of categories or elements among the multiple union information be placed at the beginning of the order (1st). At this time, the determination unit 140 may change the order according to the weight or priority assigned to the categories.

[0033] Furthermore, the decision unit 140 may, from among the content information whose order has not been determined, identify the number of categories or elements included in the common set for each of the multiple unions of content information relating to the immediately preceding (for example, the i-th) content information whose order has been determined (this may also be called "one of the content information"), and determine the other content information corresponding to the common set with the largest number of categories or elements as the (i+1)th item in the order. In this case, if there are multiple candidates, the decision unit 140 may leave them in no particular order, or it may order the unordered candidates according to the weight or priority assigned to the categories.

[0034] Furthermore, if there is union information where the intersection is an empty set, the decision unit 140 may order the content information within the remaining union information.

[0035] Furthermore, if the union information in the determination unit 140 has an empty set as its common set, the information processing device 100 may relax the classification conditions of the set unit 130. By relaxing the classification conditions, it is expected that elements will be more likely to be included in the common set, thereby reducing the amount of union information in the determination unit 140 that has an empty set as its common set.

[0036] The configuration of the information processing apparatus according to the embodiment has been described above. Next, the operation of the information processing apparatus according to the embodiment will be described.

[0037] [Operation] Figure 4 is a flowchart showing an example of the operation of the sequence presentation process in the information processing device according to the embodiment. The sequence presentation process is a process that assigns a sequence to multiple content pieces according to the overlap of content between them, and presents the result of the sequence (sequence information) to the user. The flowchart in Figure 4 starts, for example, when the user executes the sequence information presentation process.

[0038] (Step S110) The acquisition unit 110 acquires multiple pieces of content information related to the target content.

[0039] (Step S120) The extraction unit 120 extracts the content contained in each of the multiple content pieces based on the extraction conditions relating to the target content.

[0040] (Step S130) The extraction unit 120 generates a set of elements related to the content as extracted information for each of the multiple content pieces. Thus, the extraction unit 120 generates multiple pieces of extracted information corresponding to each of the multiple content pieces.

[0041] (Step S140) For every combination of two different extracted pieces of information from the multiple extracted pieces of information, the aggregation unit 130 classifies the multiple elements contained in the two extracted pieces of information into multiple categories based on classification information relating to the target content.

[0042] (Step S150) The aggregation unit 130 generates union information for the multiple classified categories, which consists of a common set of categories containing elements from each of the two extracted pieces of information, and a difference set of categories containing elements from either of the two extracted pieces of information. Thus, the aggregation unit 130 generates multiple union information corresponding to all combinations of two different extracted pieces of information from the multiple extracted pieces of information.

[0043] (Step S160) The determination unit 140 determines the ordering of multiple content information based on the union information of multiple content information and generates order information indicating the order of the multiple content information. The information processing device 100 presents the generated order information to the user. After step S160, the flowchart in Figure 4 ends.

[0044] To briefly outline the order presentation process in Figure 4, the information processing device 100 extracts content from each of the multiple content pieces related to the target content based on extraction conditions for the target content, and generates a set of elements related to the content as extracted information. For every combination of two different extracted pieces of information corresponding to each of the multiple content pieces, the device classifies the multiple elements contained in the two extracted pieces of information into multiple categories based on classification conditions for the target content. For each of the multiple categories, the device generates union information consisting of a common set, which is the set of categories containing elements contained in each of the two extracted pieces of information, and a difference set, which is the set of categories containing elements contained in either of the two extracted pieces of information. Based on the generated union information corresponding to all combinations of the multiple extracted pieces of information, the device determines the ordering of the multiple content pieces and generates order information indicating the order of the multiple content pieces.

[0045] The operation of the information processing device according to the embodiment has been described above. Hereafter, four specific examples of the embodiment will be described. The first specific example is one in which the target content is "programming teaching materials". The second specific example is one in which the target content is "operation manual". The third specific example is one in which the target content is "visual programming exercises". The fourth specific example is one in which the target content is "books of any genre".

[0046] [First Specific Example] The first specific example assumes presenting a usage order that allows the user to efficiently learn from multiple programming learning materials. In the first specific example, the extraction criteria are the text data contained in the programming learning materials, and the classification criteria are programming-related terms.

[0047] The acquisition unit 110 acquires multiple content information M related to programming teaching materials. １ ~M Ｎ This retrieves the content information, which includes text data such as course outlines, table of contents, and main text.

[0048] The extraction unit 120 extracts multiple content information M １ ~M Ｎ For each item, content is extracted from the content information based on the extraction criteria. The extracted content consists of programming terms (words). The process by which the extraction unit 120 extracts words may also be called the "word extraction process." The extraction unit 120 may also generate an "extracted word list" which is a list of the extracted words. The extraction unit 120 may also obtain the frequency of occurrence of the extracted words.

[0049] Next, the extraction unit 120 extracts multiple content information M １ ~M Ｎ For each case, the extracted word list and the keyword list are compared, and words with a similarity between them equal to or greater than an arbitrary threshold α are identified and extracted from the extracted word list. The extraction unit 120 converts the extracted words into words in the keyword list and generates a set of converted words as extracted information. The keyword list is a list of words created in advance. The process by which the extraction unit 120 converts and extracts identified words is an abstraction process that unifies variations in spelling, and may therefore be called an "abstraction process." In addition, the words included in the extracted information may be called "elements." Therefore, the extracted information can also be said to be a set of multiple elements.

[0050] The similarity between words can be obtained, for example, by vectorizing the words using a method (e.g., BERT) as described in Reference 1 (Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova, "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", arXiv:1810.04805v2[cs.CL], May 24, 2019), and then calculating the similarity between the vectors, such as cosine similarity.

[0051] Based on the above, the extraction unit 120 extracts multiple content information M １ ~M Ｎ By performing word extraction and abstraction processing on each of them, multiple content information M １ ~M Ｎ Multiple extracted pieces of information B corresponding to each １ ~B Ｎ Each of these will be generated.

[0052] The aggregation unit 130 collects multiple extracted pieces of information B. １ ~B Ｎ All combinations of two different extracted pieces of information (B ｋ , B ｌ For each of the following cases (where k = 1 to N, l = 1 to N, and k < l), the multiple elements contained in the two extracted pieces of information are classified into multiple categories based on the classification criteria for programming educational materials.

[0053] Specifically, the manifold 130 is a combination (B ｋ , B ｌ Each element of ) is classified into multiple categories according to its similarity to the words included in the classification criteria. After classifying into multiple categories, the aggregation unit 130 extracts the information B ｋ and extracted information B ｌ The category containing the elements included in each is the common category C. ｋ，ｌ，ｏ (where o = 1 to 0), and extracted information B ｋ A subcategory D is a category that contains elements that are found only in that category. ｋ，ｌ，ｐ(However, p = 1 to P), and extracted information B ｌ A subcategory D is a category that contains elements that are found only in that category. ｌ，ｋ，ｑ (However, q = 1 to Q). Furthermore, weights or priorities based on specific elements, the frequency of element occurrences, etc., may be assigned to any category.

[0054] The collection unit 130 consists of multiple classified categories (common category C ｋ，ｌ，ｏ Partial Category D ｋ，ｌ，ｐ , and partial category D ｌ，ｋ，ｑ Regarding ), the categories (common category C) that have elements included in each of the two extracted pieces of information. ｋ，ｌ，ｏ The intersection B is the set of ) ｋ ∩B ｌ , and a category (partial category D) that has elements included in either of the two extracted pieces of information. ｋ，ｌ，ｐ and partial category D ｌ，ｋ，ｑ The difference between two sets of (difference set B) ｋ \B ｌ and set difference B ｌ \B ｋ The union B consists of ) ｋ ∪B ｌ This is generated as union information. The series of processes in the set unit 130 described above may also be called "set processing".

[0055] Based on the above, the manifold 130 is a combination (B ｋ , B ｌ ) By performing set processing on each of them, the combination (B ｋ , B ｌ ) Multiple union information corresponding to each (union B １ ∪B ２ ,..., union B Ｎ－１ ∪B Ｎ ) generates.

[0056] The decision unit 140 determines the union information of multiple sets (union B １ ∪B ２ ,..., union B Ｎ－１ ∪B Ｎ Based on this, multiple content information M １ ~M Ｎ Determine the ordering of multiple content information M １to M Ｎ generates order information indicating the order for the above.

[0057] Specifically, the determining unit 140, for each of a plurality of union information pieces (union B １ ∪B ２ , ..., union B Ｎ－１ ∪B Ｎ ), identifies the number of categories (n(B ｋ \B ｌ ) included in one difference set, or the number of elements, and determines the content information M corresponding to the difference set B ｋ \B ｌ ) with the smallest number among the plurality of union information pieces (union B １ ∪B ２ , ..., union B Ｎ－１ ∪B Ｎ ) as the top (first) position in the order. At this time, the determining unit 140 may change the order according to the weight or priority assigned to the category. For example, the determining unit 140 may preferentially select a category assigned with "priority: high". ｋ \B ｌ corresponding to the difference set B ｋ at the top (first) of the order. At this time, the determining unit 140 may change the order according to the weight or priority assigned to the category. For example, the determining unit 140 may preferentially select a category assigned with "priority: high".

[0058] Next, the determining unit 140 may identify, from among pieces of content information whose order has not been determined, the number of categories or the number of elements included in the intersection for each of a plurality of union information pieces related to the immediately preceding (e.g., i-th) content information whose order has been determined, and determine the other piece of content information corresponding to the intersection with the largest number as the (i+1)-th position in the order. At this time, if there are a plurality of candidates, the determining unit 140 may leave them unordered, or may order the unordered candidates according to the weight or priority assigned to the categories. Note that the series of processes of the determining unit 140 described above may be referred to as "order determination processing".

[0059] Hereinafter, the processing by the extracting unit 120, the processing by the aggregating unit 130, and the processing by the determining unit 140 in the first specific example of the embodiment will be described using FIG. 5 to FIG. 7. In the following, three pieces of content information (content information M １ , content information M ２ , and content information M ３) is described.

[0060] FIG. 5 is an explanatory diagram of word extraction processing and abstraction processing in the first specific example of the embodiment. In FIG. 5, content information M １ to extraction information B １ is shown.

[0061] The extraction unit 120 generates an extracted word list 511 by executing word extraction processing 510 on content information 501 which is content information M １ . The content information 501 includes, for example, text data such as teaching material summaries, table of contents, and main text. The extracted word list 511 includes words extracted from the content information 501. The keyword list KL includes pre-created words related to programming.

[0062] The extraction unit 120 generates extraction information 521 that is extraction information B １ by executing abstraction processing 520 using the extracted word list 511 and the keyword list KL. The extraction information 521 includes words that are identified from the extracted word list 511 and whose notation variations are unified by the keyword list KL.

[0063] Note that the extraction unit 120 also executes the same processing for content information M ２ and content information M ３ . Therefore, the extraction unit 120 further generates extraction information B ２ from content information M ２ , and generates extraction information B ３ from content information M ３ .

[0064] FIG. 6 is an explanatory diagram of set processing in the first specific example of the embodiment. In FIG. 6, from three pieces of extraction information (extraction information B １ , extraction information B ２ , and extraction information B ３ ) to a plurality of union information (union B １ ∪B ２ , union B １ ∪B ３ , and union B ２ ∪B ３The process for generating ) is shown.

[0065] Below, the two extracted pieces of information are referred to as Extracted Information A. １ and extracted information A ２ In that case, for example, the union A １ ∪A ２ "C" is included in １，２，ｎ : {x, y}" is extracted information A １ The element "x" contained within, and extracted information A ２ This means that the element "y" contained within is classified into the nth common category. Also, for example, "C １，２，ｎ : {(x １ , x ２ ), y}" is extracted information A １ The element "x" included in １ " and element "x ２ " and extracted information A ２ This means that the element "y" contained within is classified into the nth common category.

[0066] Also, for example, "D １，２，ｎ : {x}" is extracted information A １ This means that the element "x" contained within is classified into the nth subcategory. Also, for example, "D ２，１，ｎ : {(y １ , y ２ )}" is extracted information A ２ The element "y" included in １ " and element "y ２ This means that it is classified into the nth subcategory.

[0067] The aggregation unit 130 contains extracted information 521 and extracted information B. ２ Extracted information 522 and extracted information B ３ By performing set processing 610 on the extracted information 523, the union B is formed. １ ∪B ２ The union of 611 and the union of B １ ∪B ３ The union of 612 and the union of B ２ ∪B ３ This generates a union 613.

[0068] Union set 611 (Union set B１ ∪B ２ ) Common category "C １，２，２ : {(variable, int), str}" is extracted information 521 (extracted information B １ The elements "variable" and "int" included in ) and extracted information 522 (extracted information B ２ This means that the element "str" ​​included in ) is classified in the same category.

[0069] Figure 7 is an explanatory diagram of the order determination process in the first specific example of the embodiment. Figure 7 shows multiple union information (union B １ ∪B ２ , union B １ ∪B ３ , and union B ２ ∪B ３ The process for generating order information from the above is shown. In addition, for each of the multiple union information in Figure 7, the number of categories in the intersection and the number of categories in each of the two difference sets are shown.

[0070] The determination unit 140 generates order information by executing an order determination process 710 on the union 611, union 612, and union 613. Specifically, the determination unit 140 determines the difference set B in union 611. １ \B ２ Number of categories (n(B) １ \B ２ Since ) = 0) is the smallest, extracted information B １ Corresponding content information M １ Determine this as the first item in the sequence.

[0071] Next, the determination unit 140 determines the order of the extracted information B １ Among the unions 611 and 612 that include the set B, the intersection in union 611. １ ∩B ２ Number of categories included (n(B) １ ∩B ２ Since ) = 4) is the largest, the other extracted information B ２ Corresponding content information M ２ We will decide on this second in the order.

[0072] Finally, the determination unit 140 determines the remaining extracted information B３ Corresponding content information M ３ This is determined to be the third item in the sequence. Therefore, the sequence information includes information indicating the order of the content information (M １ →M ２ →M ３ ) is included.

[0073] [Second Specific Example] The second specific example assumes presenting a usage sequence for multiple operation manuals that allows users to efficiently understand the operations. In the second specific example, the extraction criteria are the text data and visual data (mainly "image data") contained in the operation manual, and the classification criteria are the distance between elements after the individual data has been vectorized.

[0074] The acquisition unit 110 acquires multiple content information M related to the operation manual. １ ~M Ｎ The content information is retrieved, and includes, for example, text data and image data.

[0075] The extraction unit 120 extracts multiple content information M １ ~M Ｎ For each item, content is extracted from the content information based on the classification criteria. The extracted content consists of words and images. The process by which the extraction unit 120 extracts words may be called the "word extraction process," and the process by which the extraction unit 120 extracts images may be called the "image extraction process." The extraction unit 120 may also obtain the frequency of occurrence of the extracted words.

[0076] Next, the extraction unit 120 extracts multiple content information M １ ~M Ｎ For each extracted word and image, the vectorization process is performed, and the distance between these vectors is converted into a format that allows for calculation. The extraction unit 120 generates the set of vectorized data as extracted information. The process by which the extraction unit 120 vectorizes the extracted data may also be called the "vectorization process." The vectorized data included in the extracted information may also be called "elements." Furthermore, the extraction unit 120 uses an arbitrary threshold β for the frequency of occurrence. １ You may extract only the words listed above.

[0077] Furthermore, vectorization of data containing different formats such as words and images can be performed using methods such as those described in Reference 2 (Alec Radford et al. “Learning Transferable Visual Models From Natural Language Supervision”, arXiv:2103.00020v1[cs.CV], Feb. 24, 2021) (e.g., CLIP).

[0078] Based on the above, the extraction unit 120 extracts multiple content information M １ ~M Ｎ By performing word extraction, image extraction, and vectorization processes on each of them, multiple content information M １ ~M Ｎ Multiple extracted pieces of information B corresponding to each １ ~B Ｎ Each of these will be generated.

[0079] The aggregation unit 130 collects multiple extracted pieces of information B. １ ~B Ｎ All combinations of two different extracted pieces of information (B ｋ , B ｌ For each of the following cases (where k = 1 to N, l = 1 to N, k < l), the multiple elements contained in the two extracted pieces of information are classified into multiple categories based on the classification criteria related to the operation manual.

[0080] Specifically, the manifold 130 is a combination (B ｋ , B ｌ The elements of the ) are classified into multiple categories according to the distance between the elements. For example, the aggregation unit 130 is classified into categories based on the distance between the elements if the distance between the elements is an arbitrary threshold γ １ The following combinations are classified into a common category, and all others into subcategories. The subsequent processing by the aggregation unit 130 is the same as in the first specific example, so the explanation is omitted. The series of processes by the aggregation unit 130 may also be called "aggregation processing".

[0081] Based on the above, the manifold 130 is a combination (B ｋ , B ｌ) By performing set processing on each of them, the combination (B ｋ , B ｌ ) Multiple union information corresponding to each (union B １ ∪B ２ ,..., union B Ｎ－１ ∪B Ｎ ) generates.

[0082] The decision unit 140 determines the union information of multiple sets (union B １ ∪B ２ ,..., union B Ｎ－１ ∪B Ｎ Based on this, multiple content information M １ ~M Ｎ Determine the ordering of multiple content information M １ ~M Ｎ This generates sequence information indicating the order of the items. The subsequent processing by the determination unit 140 can be the same as in the first specific example, so its explanation is omitted. The series of processes by the determination unit 140 may also be called the "sequence determination process".

[0083] The following describes the processing by the extraction unit 120, the processing by the aggregation unit 130, and the processing by the determination unit 140 in a second specific example of the embodiment, using Figures 8 to 10. In the following, three types of content information (content information M) will be used. １ Content information M ２ , and content information M ３ The ordering of ) is explained.

[0084] Figure 8 is an explanatory diagram of the word extraction process, image extraction process, and vectorization process in a second specific example of the embodiment. Figure 8 shows content information M １ Extracted information B １ The process for generating this is shown.

[0085] The extraction unit 120 extracts content information M １ By performing word extraction processing 810, image extraction processing 820, and vectorization processing 830 on the content information 801, the extracted information B １Extracted information 831 is generated. Content information 801 includes, for example, text data such as manual descriptions and image data such as manual images. Extracted information 831 includes vectorized words W and vectorized images I.

[0086] The extraction unit 120 extracts content information M ２ and content information M ３ The same process is performed for the content information M. Therefore, the extraction unit 120 further processes the content information M. ２ Extracted information B ２ Generates content information M ３ Extracted information B ３ Generates.

[0087] Figure 9 is an explanatory diagram of the set processing in a second specific example of the embodiment. Figure 9 shows three types of extracted information (extracted information B １ , extracted information B ２ , and extracted information B ３ ) from multiple union information (union B １ ∪B ２ , union B １ ∪B ３ , and union B ２ ∪B ３ The process for generating ) is shown.

[0088] The aggregation unit 130 contains extracted information 831 and extracted information B. ２ Extracted information 832 and extracted information B ３ By performing set processing 910 on the extracted information 833, the union B is formed. １ ∪B ２ The union of sets 911 and B is 911. １ ∪B ３ The union of 912 and the union of B ２ ∪B ３ This generates the union 913.

[0089] Figure 10 is an explanatory diagram of the order determination process in a second specific example of the embodiment. Figure 10 shows multiple union information (union B １ ∪B ２ , union B １ ∪B ３ , and union B２ ∪B ３ The process for generating order information from the above is shown. Note that each of the multiple union information in Figure 10 shows the number of categories in the intersection set and the number of categories in each of the two difference sets.

[0090] The determination unit 140 generates order information by executing an order determination process 1010 on the union 911, union 912, and union 913. Specifically, the determination unit 140 determines the difference set B in union 911. ２ \B １ Number of categories (n(B) ２ \B １ Since ) = 0) is the smallest, extracted information B ２ Corresponding content information M ２ Determine this as the first item in the sequence.

[0091] Next, the determination unit 140 determines the order of the extracted information B ２ Among the unions 911 and 913 that include, the intersection B in union 911. １ ∩B ２ Number of categories (n(B) １ ∩B ２ Since ) = 1) is the largest, the other extracted information B １ Corresponding content information M １ We will decide on this second in the order.

[0092] Finally, the determination unit 140 determines the remaining extracted information B ３ Corresponding content information M ３ This is determined to be the third item in the sequence. Therefore, the sequence information includes information indicating the order of the content information (M ２ →M １ →M ３ ) is included.

[0093] [Third Specific Example] The third specific example envisions presenting a learning sequence that allows users to efficiently learn multiple visual programming exercises (hereinafter sometimes simply referred to as "exercises"). In the third specific example, the extraction criteria are the visual data (mainly "image data") of the blocks in the visual programming included in the exercises, or attribute information attached to the blocks themselves (for example, elements of XML (extensible markup language)), and the classification criteria are the type of block in the visual programming.

[0094] Figure 11 is an explanatory diagram of the content in a third specific example of the embodiment. The content 1101 in Figure 11 includes, for example, the problem statement of an exercise and the model answer 1111 of the exercise. The model answer 1111 shows an image of a combination of blocks in visual programming (syntax blocks (e.g., click blocks), command blocks, output blocks (e.g., value setting blocks), etc.) (combination block image), or attribute information (such as XML elements) related to the blocks included in the combination and how they are combined. This combination block image or attribute information is used as content information. Hereafter, the combination block image or attribute information of the model answer for each of the multiple exercises will be used as content information.

[0095] The acquisition unit 110 acquires multiple content information M related to the practice problems. １ ~M Ｎ This retrieves the content information, which may include, for example, image data or attribute information.

[0096] The extraction unit 120 extracts multiple content information M １ ~M Ｎ For each, the content information is extracted based on the extraction conditions. The extracted content is either a combined block image or attribute information. The extraction unit 120 extracts the content information M from multiple content information １ ~M ＮFor each case, the extracted combination block images and attribute information are classified into individual block images and attribute information. The process by which the extraction unit 120 extracts and classifies the content may also be called the "classification process." The extraction unit 120 may also obtain the number of occurrences of the extracted blocks.

[0097] Next, the extraction unit 120 extracts multiple content information M １ ~M Ｎ For each type, the classified block images and attribute information are grouped together. Specifically, the extraction unit 120 groups multiple blocks into one group, and considers the blocks included in that group to be of the same type. The extraction unit 120 then generates a set of block types as extracted information. The process by which the extraction unit 120 groups block images and attribute information by type is an abstraction process, such as aggregating into the same type, and may therefore be called an "abstraction process." The block types included in the extracted information may also be called "elements."

[0098] Based on the above, the extraction unit 120 extracts multiple content information M １ ~M Ｎ By performing classification and abstraction processing on each of them, multiple content information M １ ~M Ｎ Multiple extracted pieces of information B corresponding to each １ ~B Ｎ Each of these will be generated.

[0099] The aggregation unit 130 collects multiple extracted pieces of information B. １ ~B Ｎ All combinations of two different extracted pieces of information (B ｋ , B ｌ ) (where k = 1 to N, l = 1 to N, k < l) For each of these, the multiple elements contained in the two extracted pieces of information are classified into multiple categories based on the classification conditions for the type of block. For example, the aggregation unit 130 classifies combinations of the same block type between elements into a common category, and classifies the rest into subcategories. The subsequent processing by the aggregation unit 130 can be the same as in the first specific example, so the explanation is omitted. The series of processes by the aggregation unit 130 may also be called "aggregation processing".

[0100] Based on the above, the manifold 130 is a combination (B ｋ , B ｌ ) By performing set processing on each of them, the combination (B ｋ , B ｌ ) Multiple union information corresponding to each (union B １ ∪B ２ ,..., union B Ｎ－１ ∪B Ｎ ) generates.

[0101] The decision unit 140 determines the union information of multiple sets (union B １ ∪B ２ ,..., union B Ｎ－１ ∪B Ｎ Based on this, multiple content information M １ ~M Ｎ Determine the ordering of multiple content information M １ ~M Ｎ This generates sequence information indicating the order of the items. The subsequent processing by the determination unit 140 can be the same as in the first specific example, so its explanation is omitted. The series of processes by the determination unit 140 may also be called the "sequence determination process".

[0102] The following describes the processing by the extraction unit 120, the processing by the aggregation unit 130, and the processing by the determination unit 140 in a third specific example of the embodiment, using Figures 12 to 14. In the following, three types of content information (content information M) will be used. １ Content information M ２ , and content information M ３ The ordering of ) is explained.

[0103] Figure 12 is an explanatory diagram of the classification process and abstraction process in a third specific example of the embodiment. Figure 12 shows content information M １ Extracted information B １ The process for generating [the image] is shown. In this example, only block images are used in the process.

[0104] The extraction unit 120 extracts content information M １Block classification information 1211 is generated by performing a classification process 1210 on the content information 1201. The block classification information 1211 includes, for example, individual block images (first block image 1211A, second block image 1211B, and third block image 1211C) classified from the combined block images contained in the content information 1201. The first block image 1211A is, for example, a click block, and the second block image 1211B and third block image 1211C are, for example, value setting blocks.

[0105] Next, the extraction unit 120 performs an abstraction process 1220 on the block classification information 1211, thereby extracting the information B １ Extracted information 1221 is generated. Extracted information 1221 includes, for example, information for "BLK1" corresponding to a click block and information for "BLK2" corresponding to a value setting block. Since both the second block image 1211B and the third block image 1211C are value setting blocks, the information for these two is aggregated in "BLK2".

[0106] The extraction unit 120 extracts content information M ２ and content information M ３ The same process is performed for the content information M. Therefore, the extraction unit 120 further processes the content information M. ２ Extracted information B ２ Generates content information M ３ Extracted information B ３ Generates.

[0107] Figure 13 is an explanatory diagram of the set processing in a third specific example of the embodiment. Figure 13 shows three extracted pieces of information (extracted information B １ , extracted information B ２ , and extracted information B ３ ) from multiple union information (union B １ ∪B ２ , union B １ ∪B ３ , and union B ２ ∪B ３ The process for generating ) is shown.

[0108] The aggregation unit 130 contains extracted information 1221 and extracted information B. ２ Extracted information 1222 and extracted information B ３ By performing set processing 1310 on the extracted information 1223, the union B is formed. １ ∪B ２ The union of sets 1311 and B is 1311. １ ∪B ３ The union of 1312 and the union of B ２ ∪B ３ This generates the union 1313.

[0109] Figure 14 is an explanatory diagram of the order determination process in a third specific example of the embodiment. Figure 14 shows multiple union information (union B １ ∪B ２ , union B １ ∪B ３ , and union B ２ ∪B ３ The process for generating order information from the above is shown. In addition, for each of the multiple union information in Figure 14, the number of categories in the intersection and the number of categories in each of the two difference sets are shown.

[0110] The determination unit 140 generates order information by executing an order determination process 1410 on the union 1311, union 1312, and union 1313. Specifically, the determination unit 140 determines the difference set B in the union 1311. １ \B ２ Number of categories (n(B) １ \B ２ Since ) = 0) is the smallest, extracted information B １ Corresponding content information M １ Determine this as the first item in the sequence.

[0111] Next, the determination unit 140 determines the order of the extracted information B １ Among the unions 1311 and 1312 that include the set B, the intersection in union 1311. １ ∩B ２ Number of categories included (n(B) １ ∩B ２ Since ) = 2) is the largest, the other extracted information B ２ Corresponding content information M２ We will decide on this second in the order.

[0112] Finally, the determination unit 140 determines the remaining extracted information B ３ Corresponding content information M ３ This is determined to be the third item in the sequence. Therefore, the sequence information includes information indicating the order of the content information (M １ →M ２ →M ３ ) is included.

[0113] [Fourth Specific Example] The fourth specific example envisions presenting a usage order that allows users to efficiently acquire knowledge from books of various genres (hereinafter simply referred to as "books"). In the fourth specific example, the extraction criteria are the text data contained in the books, and the classification criteria are specific items and the distance between elements after vectorizing the individual data.

[0114] The acquisition unit 110 acquires multiple content information M related to books. １ ~M Ｎ This retrieves content information, which includes, for example, bibliographic information (author name, publication year, genre, etc.) and text data such as the main text.

[0115] The extraction unit 120 extracts multiple content information M １ ~M Ｎ For each item, content is extracted from the content information based on the extraction criteria. The extracted content consists of words found in the main text of the book. The process by which the extraction unit 120 extracts words may also be called the "word extraction process." The extraction unit 120 may also generate an "extracted word list" which is a list of the extracted words. Furthermore, the extraction unit 120 may obtain the frequency of occurrence of the extracted words.

[0116] Next, the extraction unit 120 extracts multiple content information M １ ~M ＮFor each case, the extracted words are vectorized and converted into a format in which the distance between those vectors can be calculated. The extraction unit 120 generates extracted information as a set of data consisting of the unvectorized bibliographic information data and the vectorized data. The process by which the extraction unit 120 vectorizes the extracted words may also be called the "vectorization process." Furthermore, each data item included in the extracted information may be called an "element." In addition, the extraction unit 120 uses an arbitrary threshold β for the number of occurrences. ２ You may extract only the words listed above.

[0117] Based on the above, the extraction unit 120 extracts multiple content information M １ ~M Ｎ By performing word extraction and vectorization processing on each of them, multiple content information M １ ~M Ｎ Multiple extracted pieces of information B corresponding to each １ ~B Ｎ Each of these will be generated.

[0118] The aggregation unit 130 collects multiple extracted pieces of information B. １ ~B Ｎ All combinations of two different extracted pieces of information (B ｋ , B ｌ For each of the following cases (where k = 1 to N, l = 1 to N, k < l), the multiple elements contained in the two extracted pieces of information are classified into multiple categories based on specific items and classification conditions relating to the distance between elements after the individual data have been vectorized.

[0119] Specifically, the manifold 130 is a combination (B ｋ , B ｌ For each element of the ), the bibliographic information is classified into categories corresponding to specific items (i.e., items of bibliographic information). For example, the aggregation unit 130 classifies the items of bibliographic information into common categories or subcategories based on the degree of matching of the elements.

[0120] Furthermore, the aggregation unit 130 classifies the vectorized elements into multiple categories according to the distance between the elements. For example, the aggregation unit 130 classifies the elements into an arbitrary threshold γ ２The following combinations are classified into a common category, and all others into subcategories. The series of processes in the aggregation unit 130 may also be called "aggregation processing."

[0121] Based on the above, the manifold 130 is a combination (B ｋ , B ｌ ) By performing set processing on each of them, the combination (B ｋ , B ｌ ) Multiple union information corresponding to each (union B １ ∪B ２ ,..., union B Ｎ－１ ∪B Ｎ ) generates.

[0122] The decision unit 140 determines the union information of multiple sets (union B １ ∪B ２ ,..., union B Ｎ－１ ∪B Ｎ Based on this, multiple content information M １ ~M Ｎ Determine the ordering of multiple content information M １ ~M Ｎ Generates ordering information indicating the order of the items.

[0123] Specifically, the determination unit 140 may generate sequence information in at least one of two patterns: a first process that does not perform priority ordering, and a second process that performs priority ordering. The first process can be the same as in the first specific example, so its explanation will be omitted. In the second process, content information relating to combinations in which bibliographic information items are classified into the same category is prioritized for sequence ordering. The above series of processes by the determination unit 140 may also be called the "sequence determination process."

[0124] The following describes the processing by the extraction unit 120, the processing by the aggregation unit 130, and the processing by the determination unit 140 in the fourth specific example of the embodiment, using Figures 15 to 17. In the following, three types of content information (content information M) will be used. １ Content information M ２ , and content information M ３ The ordering of ) is explained.

[0125] Figure 15 is an explanatory diagram of the word extraction process and vectorization process in a fourth specific example of the embodiment. Figure 15 shows content information M １ Extracted information B １ The process for generating this is shown.

[0126] The extraction unit 120 extracts content information M １ A word extraction process 1510 is performed on the content information 1501 to generate an extracted word list 1511. The content information 1501 includes text data such as bibliographic information and the main text. The extracted word list 511 contains words extracted from the main text of the content information 1501.

[0127] Next, the extraction unit 120 generates word vectors by performing a vectorization process 1520 on the extracted word list 1511. The extraction unit 120 also extracts bibliographic information from the content information 1501. The extraction unit 120 then generates extracted information B, which includes the extracted bibliographic information and the generated word vectors. １ This generates extracted information 1521.

[0128] The extraction unit 120 extracts content information M ２ and content information M ３ The same process is performed for the content information M. Therefore, the extraction unit 120 further processes the content information M. ２ Extracted information B ２ Generates content information M ３ Extracted information B ３ Generates.

[0129] Figure 16 is an explanatory diagram of the set processing in a fourth specific example of the embodiment. Figure 16 shows three extracted pieces of information (extracted information B １ , extracted information B ２ , and extracted information B ３ ) from multiple union information (union B １ ∪B ２ , union B １ ∪B ３ , and union B ２ ∪B ３ The process for generating ) is shown.

[0130] The aggregation unit 130 contains extracted information 1521 and extracted information B. ２ Extracted information 1522 and extracted information B ３ By performing set processing 1610 on the extracted information 1523, the union B is formed. １ ∪B ２ The union of 1611 and the union of B １ ∪B ３ The union of 1612 and the union of B ２ ∪B ３ This generates the union 1613. Note that a specific example of the union 1613 is omitted in Figure 16.

[0131] Figure 17 is an explanatory diagram of the order determination process in a fourth specific example of the embodiment. Figure 17 shows multiple union information (union B １ ∪B ２ , union B １ ∪B ３ , and union B ２ ∪B ３ The process for generating order information from the above is shown. In addition, each of the multiple union information in Figure 17 shows the number of categories in the intersection set, the number of categories in each of the two difference sets, and the category matching information of the bibliographic information. The category matching information is information in which categories whose bibliographic information items match are designated as "same category," and categories whose bibliographic information items do not match are designated as "different category."

[0132] The determination unit 140 generates order information by executing an order determination process 1710 on the union 1611, union 1612, and union 1613. The generation of order information will be described below for each case of the first and second processes.

[0133] In the first process, the determination unit 140 determines the difference set B in the union 1612. １ \B ３ Number of categories (n(B) １ \B ３ Since ) = 3) is the smallest, extracted information B １ Corresponding content information M １ Determine this as the first item in the sequence.

[0134] Next, the determination unit 140 determines the order of the extracted information B １ Among the unions 1611 and 1612 that include the set B, the intersection in union 1612 is １ ∩B ３ Number of categories included (n(B) １ ∩B ３ Since ) = 6) is the largest, the other extracted information B ３ Corresponding content information M ３ We will decide on this second in the order.

[0135] Finally, the determination unit 140 determines the remaining extracted information B ２ Corresponding content information M ２ This is determined to be the third item in the sequence. Therefore, the sequence information includes information indicating the order of the content information (M １ →M ３ →M ２ ) is included.

[0136] In the second process, the determination unit 140 compares the two sets of categories in the difference set in ascending order, and the difference is the threshold θ １ In the following cases, the order of content information is determined based on category matching information. Specifically, the determination unit 140 determines the difference set B in the union set 1612. １ \B ３ Number of categories (n(B) １ \B ３ ) = 3) is the smallest, and the difference set B in the union 1611 １ \B ２ Number of categories (n(B) １ \B ２ Since () = 5) is the second smallest, we calculate the difference in the number of categories. At this time, the difference in the number of categories of the difference set is the threshold θ １ Assuming the following, the decision unit 140 refers to the category matching information and, for example, prioritizes the union set 1611 where the author name is in the "same category". Therefore, the decision unit 140 extracts information B １ Corresponding content information M １ Determine this as the first item in the sequence.

[0137] Next, the determination unit 140 compares the two sets of categories in the union in descending order, and the difference between them is the threshold θ. ２In the following cases, the order of content information is determined based on category matching information. Specifically, the determination unit 140 determines the order of extracted information B １ In unions 1611 and 1612, which include the specified set, the difference in the number of categories in the unions is calculated. At this time, the difference in the number of categories in the unions is a threshold θ. ２ Assuming the following, the decision unit 140 refers to the category matching information and, for example, prioritizes the union set 1611 where the author name is in the "same category". Therefore, the decision unit 140 considers the other extracted information B. ２ Corresponding content information M ２ We will decide on this second in the order.

[0138] Finally, the determination unit 140 determines the remaining extracted information B ３ Corresponding content information M ３ This is determined to be the third item in the sequence. Therefore, the sequence information includes information indicating the order of the content information (M １ →M ２ →M ３ ) is included.

[0139] [Effect] According to the above embodiment, the information processing device extracts content from each of the multiple content pieces relating to the target content based on extraction conditions relating to the target content, generates a set of elements relating to the content as extracted information, classifies the multiple elements contained in the two extracted pieces into multiple categories based on classification conditions relating to the target content for each of the two different sets of extracted pieces of information corresponding to each of the multiple content pieces, generates union information for the multiple categories, which consists of a common set of categories that have elements contained in each of the two extracted pieces of information, and a difference set of two sets of categories that have elements contained in either of the two extracted pieces of information, and determines the ordering of the multiple content pieces based on the generated union information corresponding to each of the multiple sets of extracted pieces of information, and generates ordering information indicating the order of the multiple content pieces of information.

[0140] Therefore, by utilizing a common set of elements that constitute the content information, the information processing device can consider the overlap between different pieces of content information and perform ordering accordingly. Because the content information before and after the ordered pieces of content information is related to each other, the information processing device can present a content usage order that allows for efficient knowledge acquisition.

[0141] Furthermore, if the extracted content includes text data to be processed, the information processing device may generate elements by performing at least one of the following processes on the text data: filtering using keywords related to the target content, filtering based on the number of occurrences of the same text data included in the content information, vectorization, and abstraction based on the similarity between text data. If the extracted content includes visual data to be processed, the information processing device may generate elements by performing at least one of the following processes on the visual data: classification, vectorization, and abstraction based on the similarity between visual data.

[0142] This allows the information processing device to present a content usage sequence that enables efficient knowledge acquisition, regardless of the type or field of the content information.

[0143] Furthermore, the information processing device may generate union information by assigning weights or priorities to categories, and may also generate order information based on the weights or priorities.

[0144] This allows the information processing device to present a content usage sequence that enables efficient knowledge acquisition, even with respect to specific content information.

[0145] <Other Embodiments> The functions of the elements disclosed herein may be implemented using circuits or processing circuits that include general-purpose processors, special-purpose processors, integrated circuits, ASICs (Application Specific Integrated Circuits), FPGAs (Field Programmable Gate Arrays), other conventional general-purpose circuits, and / or combinations thereof that are programmed using one or more programs stored in one or more memories, or otherwise configured to perform the disclosed functions. A processor is considered a processing circuit or circuit because it includes transistors and other circuits. A processor may be a programmed processor that executes programs stored in memory.

[0146] In this disclosure, a circuit, unit, or means is hardware that performs the enumerated functions, or hardware programmed to perform the enumerated functions. Hardware may be any hardware disclosed herein that is programmed or configured to perform the enumerated functions.

[0147] Furthermore, the computer instructions contained within the computer program provide logic and routines that enable hardware (e.g., processing circuits or circuits) to perform the methods disclosed herein. This computer program can be implemented in commonly known forms, such as computer-readable storage media, computer program products, memory devices, recording media such as CD-ROMs and DVDs, and / or memory for FPGAs and ASICs.

[0148] It should be noted that the present invention is not limited to the embodiments described above, and can be modified in various ways during implementation without departing from its essence. Furthermore, each embodiment may be combined as appropriate, and in that case, the combined effects can be obtained. Moreover, the above embodiments include various inventions, and various inventions can be extracted by selecting combinations from the multiple components disclosed. For example, if the problem can be solved and effects can be obtained even if some components are deleted from all the components shown in the embodiment, then the configuration with these components deleted can be extracted as an invention.

[0149] 100... Information processing device 101... Control circuit 102... Storage 103... Communication module 104... Interface 105... Drive 106... Storage medium 110... Acquisition unit 120... Extraction unit 130... Aggregation unit 140... Determination unit 501, 801, 1201, 1501... Content information 510, 810, 1510... Word extraction processing 511, 1511... Extracted word list 520, 1220... Abstraction processing 521, 522, 523, 831, 832, 833, 1221, 1222, 1223, 1521, 1522, 1523... Extracted information 610, 910, 1310, 1610... Aggregation processing 611, 612, 613, 911, 912, 913, 1311, 1312, 1313, 1611, 1612, 1613... Union 710, 1010, 1410, 1710... Order determination process 820... Image extraction process 830, 1520... Vectorization process 1101... Content 1111... Model answer 1210... Classification process 1211... Block classification information 1211A... First block image 1211B... Second block image 1211C... Third block image

Claims

1. An information processing device comprising: an extraction unit that extracts content from each of multiple pieces of content information relating to the target content based on extraction conditions relating to the target content and generates a set of elements relating to the content as extracted information; an aggregation unit that, for each combination of two different pieces of extracted information corresponding to each of the multiple pieces of content information, classifies the multiple elements contained in the two pieces of extracted information into multiple categories based on classification conditions relating to the target content, and generates a union of the multiple categories, which consists of a common set that is a set of categories having elements contained in each of the two pieces of extracted information, and a difference set that is a set of categories having elements contained in either of the two pieces of extracted information, as union information; and a determination unit that determines the ordering of the multiple pieces of content information based on the generated union information corresponding to each of the multiple combinations relating to the multiple pieces of extracted information and generates order information indicating the order of the multiple pieces of content information.

2. The information processing apparatus according to claim 1, wherein the extraction unit generates the elements by performing at least one of the following on the text data: a filtering process using keywords related to the target content, a filtering process based on the number of occurrences of the same text data included in the content information, a vectorization process, and an abstraction process based on the similarity between text data, if the extracted content includes visual data to be processed, if the extraction unit generates the elements by performing at least one of the following on the visual data: a classification process, a vectorization process, and an abstraction process based on the similarity between visual data.

3. The information processing apparatus according to claim 1 or claim 2, wherein the aggregation unit generates union information with weights or priorities assigned to the categories, and the determination unit further generates order information based on the weights or priorities.

4. An information processing method comprising: a computer extracting content from each of multiple pieces of content information relating to the target content based on extraction conditions relating to the target content, and generating a set of elements relating to the content as extracted information; classifying the multiple elements contained in the two extracted pieces of information into multiple categories based on classification conditions relating to the target content for each of the two different combinations of the extracted pieces of information corresponding to each of the multiple pieces of content information, and generating a union of the multiple categories consisting of a common set which is a set of categories having elements contained in each of the two extracted pieces of information, and a difference set which is a set of categories having elements contained in either of the two extracted pieces of information, as union information; and determining the ordering of the multiple pieces of content information based on the generated union information corresponding to each of the multiple combinations relating to the multiple pieces of content information, and generating order information indicating the order of the multiple pieces of content information.

Images