Patent analysis support device, program, and patent analysis support method

Abstract

To obtain a patent analysis support device that can present a wide range of information related to patent documents to the user. [Solution] The patent analysis support device 10 according to this disclosure includes a reception unit 11 that receives input from the user of a central phrase to be used as the focus of the analysis of a patent document, an extraction unit 14 that extracts supplementary phrases related to the central phrase from each of the multiple patent documents to be analyzed, and a display unit 15 that outputs the central phrase and the supplementary phrase.

Description

Technical Field

[0001] The present disclosure relates to a patent analysis support device, a program, and a patent analysis support method that support the utilization of patents.

Background Art

[0002] Analysis of patent documents is performed for various reasons, such as effectively utilizing owned patents and developing new ideas from existing technologies. Patent Document 1 discloses a patent utilization support device that extracts a plurality of patent documents highly relevant to a target patent, which is the patent to be utilized, and estimates a holder or inventor with high needs for the target patent based on information on the holders or inventors of the extracted patent documents, thereby generating utilization information for utilizing the target patent.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, since the patent utilization support device described in Patent Document 1 generates utilization information based on the holders or inventors, it can only generate information within an extremely limited range. For this reason, it cannot present a wide range of information unrestricted by holders or inventors to users.

[0005] The present disclosure has been made in view of the above, and an object thereof is to obtain a patent analysis support device capable of presenting a wide range of information related to patent documents to users.

Means for Solving the Problems

[0006] To solve the above-mentioned problems and achieve the objectives, the patent analysis support device according to this disclosure comprises: a reception unit that receives input from a user of central phrases to be used as the focus of the analysis of patent documents; an extraction unit that extracts supplementary phrases related to the central phrases from each of the multiple patent documents to be analyzed; and an output unit that outputs the central phrases and supplementary phrases. [Effects of the Invention]

[0007] The patent analysis support device described herein has the effect of being able to present a wide range of information related to patent documents to the user. [Brief explanation of the drawing]

[0008] [Figure 1] This figure shows an example configuration of the patent analysis support device according to Embodiment 1. [Figure 2] A flowchart showing an example of the procedure for the patent analysis support processing performed by the patent analysis support device of Embodiment 1. [Figure 3] This figure shows an example configuration of the extraction unit in Embodiment 1 when machine learning is used. [Figure 4] This figure shows an example of a display screen for the central text and supplementary text of Embodiment 1. [Figure 5] This figure shows another example of the display screen for the central text and supplementary text of Embodiment 1. [Figure 6] This figure shows an example configuration of a computer system that realizes the patent analysis support device of Embodiment 1. [Figure 7] This figure shows an example of the display screen for the central text, the first supplementary text, and the second supplementary text of Embodiment 2. [Figure 8] This figure shows an example configuration of the patent analysis support device according to Embodiment 3. [Figure 9] A diagram showing an example of a network diagram of Embodiment 3. [Figure 10] This figure shows an example of a display screen for the central text and supplementary text of Embodiment 3. [Figure 11] A diagram showing an example of a heatmap in Embodiment 3. [Figure 12]This figure shows an example of a display screen for the central text and supplementary text of Embodiment 5. [Modes for carrying out the invention]

[0009] The patent analysis support apparatus, program, and patent analysis support method according to the embodiment will be described in detail below with reference to the drawings.

[0010] Embodiment 1. Figure 1 shows an example of the configuration of a patent analysis support device 10 according to Embodiment 1. The patent analysis support device 10 of this embodiment includes a reception unit 11, a document acquisition unit 12, an information storage unit 13, an extraction unit 14, and a display unit 15.

[0011] The patent analysis support device 10 of this embodiment extracts supplementary text, which is text connected to a central phrase specified by the user, from two or more patent documents, and outputs the central phrase and the supplementary text. The central phrase is the central phrase in the analysis of the patent document, and is, for example, a word that indicates a technical element that the user wants to focus on, but is not limited to this. The central phrase is specified, for example, on a word-by-word basis, but is not limited to this, and may be specified as a word and its modifiers, such as "high-resolution camera," or it may be specified as a sentence, and the unit of the central phrase is not limited to a word. Below, an example in which the central phrase is input as a word will be described, but as mentioned above, the central phrase is not limited to a word.

[0012] The patent analysis support device 10 outputs, for example, by displaying the central text and supplementary text, but is not limited to this. It may also output the central text and supplementary text by transmitting them to another device, such as a user-operable terminal device or a user-visible display device. When the central text and supplementary text are transmitted to another device, the device that receives the central text and supplementary text displays them.

[0013] Patent documents include, for example, patent laid-open publications, patent publications (patent gazettes), and international publications of PCT (Patent Cooperation Treaty) applications, but may also include registered utility model gazettes and the like. The patent documents are not limited to Japanese-language ones, and may include patent documents issued overseas. In addition, the analysis targets of the patent analysis support device 10 include patent documents, and may further include non-patent documents such as academic papers. In the following, an example in which the documents to be analyzed are patent documents will be described, but the present invention is not limited to this. For example, one of the two documents from which supplementary phrases are extracted may be a patent document and the other may be a non-patent document. Hereinafter, patent documents are also simply referred to as documents.

[0014] The reception unit 11 receives input from a user. The reception unit 11 receives, for example, an input of a core phrase from a user. The reception unit 11 may also receive an input of information indicating a document to be analyzed (hereinafter also referred to as a document to be analyzed). The reception unit 11 receives an input by receiving a user's operation, for example, but is not limited to this. Another device such as a terminal device operable by the user may receive the input from the user and transmit the received information (input information) to the patent analysis support device 10. In this case, the reception unit 11 receives the input from the user by receiving the input information from the other device. The information indicating a document may be, for example, the number of a patent publication, or may be specified by a combination of the filing date, the title of the invention, the applicant, and the like.

[0015] The document acquisition unit 12 acquires documents to be analyzed. For example, it acquires documents via the Internet or the like from the database of an institution that provides external electronic data of patent documents, and stores the acquired electronic data of the documents (hereinafter, the electronic data of the documents are also simply referred to as documents) in the information storage unit 13. Note that the information storage unit 13 may have a function as a database of patent documents, and the electronic data of patent documents may be accumulated inside the information storage unit 13. In this case, the patent analysis support device 10 does not need to include the document acquisition unit 12. Also when acquiring non-patent documents, the document acquisition unit 12 similarly acquires non-patent documents from the database of an external institution or the like that provides electronic data of non-patent documents, and stores the acquired non-patent documents in the information storage unit 13.

[0016] The information storage unit 13 stores the documents acquired by the document acquisition unit 12. In addition, the information storage unit 13 may store the core term received by the reception unit 11 and the auxiliary terms extracted by the extraction unit 14.

[0017] The extraction unit 14 extracts auxiliary terms associated with the core term from each of the plurality of documents to be analyzed. Specifically, the extraction unit 14 extracts auxiliary terms associated with the core term received by the reception unit 11 from each of the plurality of documents stored in the information storage unit 13, and outputs the core term and the auxiliary terms to the display unit 15. The auxiliary terms may be words that appear many times in the same sentence as the core term (words with high co-occurrence frequency in the co-occurrence matrix), words that have a dependency relationship through natural language processing, words that belong to the same topic when a sentence containing the core term is analyzed using a topic model, or words for which the number of words from the core term to that word falls within a predetermined number of words a large number of times. Further, the auxiliary terms may be words represented by vectors whose inner product with the vector of the auxiliary term is equal to or greater than a certain value when words are represented as vectors through vectorization in natural language processing. The auxiliary terms are not limited to these, and may be any words associated with the core term. The extraction unit 14 stores the core term and the auxiliary terms for each document in the information storage unit 13.

[0018] The display unit 15 is an example of an output unit that outputs central text and supplementary text, and displays the central text and supplementary text for each document received from the extraction unit 14. Alternatively, the extraction unit 14 may store the central text and supplementary text for each document in the information storage unit 13, and the display unit 15 may read the central text and supplementary text for each document from the information storage unit 13 and display the read central text and supplementary text for each document. The display unit 15 may also display the central text and supplementary text as text in a table format, or it may perform processing such as making the relationship between the central text and supplementary text easier to understand before displaying them. Alternatively, instead of the display unit 15 performing this display processing, a display data generation unit (not shown) may perform the processing to generate display data that shows the display screen, and the display unit 15 may perform the display based on the display data.

[0019] Figure 1 shows an example in which the patent analysis support device 10 displays central text and supplementary text. However, as described above, the patent analysis support device 10 may also output central text and supplementary text by transmitting them to another device. In this case, the patent analysis support device 10 includes a transmission unit, which is an example of an output unit. The output unit, like the display unit 15, may generate display data that shows a display screen by performing processing such as making the relationship between central text and supplementary text easier to understand, and transmit the display data to another device. In this case, the generation of the display data may be performed by the output unit, the extraction unit 14, or a display data generation unit (not shown).

[0020] Next, the operation of this embodiment will be described. Figure 2 is a flowchart showing an example of the procedure for the patent analysis support processing performed by the patent analysis support device 10 of this embodiment. The patent analysis support device 10 receives input of information indicating the literature to be analyzed (step S1). Specifically, the reception unit 11 receives input of information indicating the literature to be analyzed (hereinafter also referred to as document identification information) from the user and outputs the document identification information to the document acquisition unit 12.

[0021] Note that there may be two or more documents to be analyzed. Below, we will explain using the case where there are two documents to be analyzed as an example. These two documents can be determined in any way, but for example, one of the two documents may be a patent document that the user owns and is considering using (hereinafter also referred to as the reference document), and the other may be a document in a technical field that is a candidate for application of the reference document. The technical field that is a candidate for application may be a different technical field from that of the reference document. For example, when considering the application of a technical element contained in the reference document to other technical fields, by designating that technical element as the central phrase and designating a document in a different technical field from that of the reference document as the subject of analysis, it is possible to consider the application of the technical element to a variety of technical fields. Note that there may be multiple reference documents.

[0022] The patent analysis support device 10 receives input of the central phrase (step S2). Specifically, the reception unit 11 receives input of the central phrase from the user and outputs the central phrase to the extraction unit 14.

[0023] The patent analysis support device 10 acquires the documents to be analyzed (step S3). Specifically, the document acquisition unit 12 acquires the documents to be analyzed from an external database based on the document identification information, as described above, and stores the acquired documents in the information storage unit 13. Alternatively, as described above, the information storage unit 13 may also function as a patent document database. In this case, for example, the extraction unit 14 acquires the documents to be analyzed by reading them from the information storage unit 13 based on the document identification information, so step S3 does not need to be performed.

[0024] The patent analysis support device 10 extracts supplementary texts related to the central phrase from the documents to be analyzed (step S4). Specifically, the extraction unit 14 reads the documents to be analyzed from the information storage unit 13 and extracts supplementary texts related to the central phrase from each document.

[0025] Here, we will describe the details of the processing of the extraction unit 14. The extraction unit 14 may, for example, use a rule-based approach to extract supplementary text from the document based on the central text. This rule may be stored in the information storage unit 13 and read by the extraction unit 14 from the information storage unit 13, or it may be held by the extraction unit 14.

[0026] When supplementary text is extracted using a rule-based approach, this rule is a rule for determining whether a word is supplementary text. For example, it could be a rule that the number of times a word is used in the same sentence as the main word exceeds a threshold, or a rule that a dependency relationship exists through natural language processing, such as the relationship between a subject and an object. Alternatively, this rule could be that the number of words between the main word and the word falls within a specified number of words exceeds a threshold, or that when words are represented as vectors through natural language processing, the dot product with the vector of the supplementary text is greater than or equal to a certain value. The rules for determining whether a word is supplementary text are not limited to these examples. The extraction unit 14 has processing functions according to the rules. For example, if a rule requiring natural language processing is used as the rule for determining whether a word is supplementary text, the extraction unit 14 has a function to perform natural language processing.

[0027] Furthermore, the extraction unit 14 may extract supplementary text using a trained model generated by machine learning. This trained model may be a trained model that has been trained to infer related supplementary text, or it may be a so-called generative AI (Artificial Intelligence) model such as a large-scale language model. The generative AI is a functional unit that generates output, i.e., an answer, using the trained model based on instructional sentences or questions called prompts.

[0028] Figure 3 shows an example configuration of the extraction unit 14 in this embodiment when machine learning is used. The extraction unit 14 shown in Figure 3 comprises a learning unit 141, a model storage unit 142, and an inference unit 143.

[0029] The learning unit 141 generates a trained model using machine learning and stores the generated trained model in the model storage unit 142. The model storage unit 142 stores the trained model. The inference unit 143 obtains output from the trained model by inputting the literature to be analyzed and the main phrases into the trained model stored in the model storage unit 142. The output of the trained model includes supplementary phrases.

[0030] For example, if the trained model is a trained model for inferring related supplementary text generated by supervised machine learning, the learning unit 141 generates the trained model by supervised learning using multiple training datasets that include, for example, literature and supplementary text as input data (explanatory variables) and corresponding ground truth supplementary text data. The ground truth data can be, for example, provided by a person, but is not limited to that. Note that the literature used as input data during training is not limited to patent documents, but can be any text. The supervised learning algorithm can, for example, be a neural network model, a CNN (Convolutional Neural Network) model, an RNN (Recurrent Neural Network), a support vector machine, etc., but is not limited to these.

[0031] For example, if the trained model is a generative AI model, the learning unit 141 generates the trained model by performing learning according to the generative AI algorithm. Generative AI models can include GAN (Generative Adversarial Networks) models, diffusion models, Transformer models, LLM (Large Language Model), VLM (Visual Language Model), BERT (Bidirectional Encoder Representations from Transformers), GPT (Generative Pre-trained Transformer), CLIP (Contrastive Language Image Pre-training), and may also use neural networks, CNNs, RNNs, or VAEs (Variational Autoencoders). Note that the above models are not mutually exclusive; for example, LLM, VLM, BERT, and GPT are included in the Transformer model.

[0032] When a generative AI is used, the inference unit 143 takes the literature to be analyzed and the central phrase as input, along with a prompt that instructs the unit to extract phrases related to the central phrase from the literature, such as "Please extract phrases related to the central phrase from the literature." The prompt may also include information indicating what kind of phrases are related.

[0033] In the example shown in Figure 3, the extraction unit 14 includes a learning unit 141. However, the system is not limited to this, and an information processing device that generates a trained model may be provided separately from the patent analysis support device 10, and the information processing device may generate the trained model. In this case, the trained model generated by the information processing device is stored in the model storage unit 142 of the extraction unit 14. Alternatively, the extraction unit 14 may not store the trained model, and the information processing device may store it. That is, the information processing device may include a model storage unit 142 and an inference unit 143. In this case, the extraction unit 14 generates data to be input to the trained model, transmits the generated data to the information processing device, and receives the output of the trained model from the information processing device. This data exchange may be performed, for example, via a transmitting / receiving unit (not shown), or the extraction unit 14 may have the functions of a transmitting / receiving unit. When the information processing device stores the trained model, the information processing device and the extraction unit 14 become an extraction unit that extracts supplementary text from documents.

[0034] Furthermore, a multimodal generation AI may be used as the generation AI, in which case not only supplementary text but also drawings connected to the central text may be extracted as supplementary drawings. In addition, a central drawing may be specified in addition to the central text, and drawings connected to the central drawing may be extracted as supplementary drawings. Connected drawings are, for example, drawings that contain the central text or figures that show technical elements corresponding to the central text, but may also include drawings that contain figures that show technical elements corresponding to the supplementary text. Connected drawings are not limited to these. Also, if a central drawing is specified, drawings similar to the central drawing may be extracted as supplementary drawings.

[0035] Returning to the explanation of Figure 2, after step S4, the patent analysis support device 10 displays the central text and supplementary text (step S5) and terminates the process. Specifically, the display unit 15 displays the central text and supplementary text. Note that Figure 2 is an example, and any process that yields equivalent results is acceptable; the specific order and content of the process are not limited to the example shown in Figure 2. For example, the order of steps S1 and S2 may be reversed.

[0036] The display unit 15 may display the central text and supplementary text in any format, but for example, the central text and supplementary text may be displayed in a format that makes it easy to visualize the connection between the central text and supplementary text for each document.

[0037] Figure 4 shows an example of a display screen for the central text and supplementary text in this embodiment. In the example shown in Figure 4, the documents to be analyzed are Document 1, which is an example of the first document, and Document 2, which is an example of the second document. In the example shown in Figure 4, the supplementary text extracted from Document 1 is displayed radially on the left side, connected to the central text by lines, and the supplementary text extracted from Document 2 is displayed radially on the right side, connected to the central text by lines. Note that the generation of display data for displaying the central text and supplementary text display screen may be performed using a generation AI.

[0038] Figure 5 shows another example of the display screen for the central text and supplementary text of this embodiment. In the example shown in Figure 5, the documents to be analyzed are Document 1, which is an example of the first document, and Document 3, which is an example of the second document. In the example shown in Figure 5, the supplementary text extracted from Document 1 is displayed on the left side, connected to the central text by a line, and the supplementary text extracted from Document 3 is displayed on the right side, connected to the central text by a line. In Figures 4 and 5, words enclosed in solid lines are the central text, and words enclosed in dashed lines are the supplementary text. In Figures 4 and 5, the sections labeled Document 1, Document 2, and Document 3 at the top display, for example, the document identification information for each document. This section may also display information such as the applicant of each document, bibliographic information such as the patent classification FI (File Index), F-term (File Forming Term), IPC (International Patent Classification), and examination status.

[0039] The user may decide how to determine the literature to be analyzed, but for example, as mentioned above, suppose there is literature that discloses the technical elements being considered for use. In that case, the user designates that literature as the reference literature and specifies the reference literature and literature in a different technical field from that reference literature as the literature to be analyzed. For example, in the examples in Figures 4 and 5, literature 1 is the reference literature, and literature 2 and 3 are designated as literature in a different technical field from that reference literature. This allows the user to check the supplementary text connected to the central text and to consider what technologies the technical elements being considered for use can be applied to. In this way, the patent document support device 10 also functions as a patent utilization support device.

[0040] For example, in Figures 4 and 5, thermal images are specified as the central text. In the example shown in Figure 4, the supplementary text extracted from references 1 and 2 includes terms such as human perceived temperature and food management temperature, and air conditioning control and refrigeration control, with supplementary text having similar meanings from each of the two references. In this case, for example, the user may determine that the technical elements described in reference 1 can potentially be applied to the technical field of reference 2 because the meanings of the supplementary texts are similar.

[0041] On the other hand, in the example shown in Figure 5, the supplementary text extracted from Reference 3 includes terms such as the trajectory and targeting of the flying object. The meanings of the supplementary text extracted from Reference 1 and Reference 3 are completely different, and there is no supplementary text with a similar meaning. In such a case, the user may conclude that it is difficult to adapt the technical elements described in Reference 1 to the technical field of Reference 2. It should be noted that the user's judgment described above is illustrative, and the purpose of using the displayed supplementary text is not limited to the example above.

[0042] Furthermore, for example, if documents 1-3 are to be analyzed, the patent analysis support device 10 may output the supplementary text of documents 1-3 along with the central text. For example, the display portion for document 3 on the right side of Figure 5 may be displayed to the right of the display portion for document 2 in Figure 4, and the supplementary text of three or more documents may be displayed along with the central text. Thus, the number of documents to be analyzed and output simultaneously does not need to be limited to two, and the extraction unit 14 only needs to extract the supplementary text that is connected to the central text from each of the multiple documents. Also, Figures 4 and 5 are illustrative examples, and the method of displaying the central text and supplementary text is not limited to these examples; for example, the supplementary text for each central text may be displayed in list format.

[0043] Next, the hardware configuration of the patent analysis support device 10 of this embodiment will be described. In the patent analysis support device 10 of this embodiment shown in Figure 1, the computer system functions as the patent analysis support device 10 when a program, which is a computer program describing the processing in the patent analysis support device 10, is executed on the computer system. Figure 6 is a diagram showing an example of the configuration of a computer system that realizes the patent analysis support device 10 of this embodiment. As shown in Figure 6, this computer system comprises a control unit 101, an input unit 102, a storage unit 103, a display unit 104, a communication unit 105, and an audio output unit 106, which are connected via a system bus 107. The control unit 101 and the storage unit 103 constitute a processing circuit.

[0044] In Figure 6, the control unit 101 is a processor such as a CPU (Central Processing Unit) and executes a program describing the processing in the patent analysis support device 10 of this embodiment. Note that a part of the control unit 101 may be implemented by dedicated hardware such as a GPU (Graphics Processing Unit) or FPGA (Field-Programmable Gate Array). The input unit 102 is a button, keyboard, mouse, touchpad, etc. The storage unit 103 includes various types of memory such as RAM (Random Access Memory) and ROM (Read Only Memory), as well as storage devices such as a hard disk, and stores the program to be executed by the control unit 101, necessary data obtained during the processing, etc. The storage unit 103 is also used as a temporary storage area for the program. The display unit 104 is a display, such as an LCD (Liquid Crystal Display). Note that the display unit 104 and the input unit 102 may be integrated and implemented as a touch panel or the like. The communication unit 105 is a receiver and transmitter that perform communication processing. The audio output unit 106 is a speaker or the like. Note that Figure 6 is just one example, and the configuration of the computer system is not limited to the example in Figure 6. For example, the computer system that implements the patent analysis support device 10 does not have to include the display unit 104 and the audio output unit 106.

[0045] Here, an example of the operation of the computer system until the program of this embodiment becomes executable will be described. In a computer system with the above configuration, for example, a computer program is installed in the storage unit 103 from a CD-ROM or DVD-ROM set in a CD (Compact Disc)-ROM drive or DVD (Digital Versatile Disc)-ROM drive (not shown). When the program is executed, the program read from the storage unit 103 is stored in the main memory area of ​​the storage unit 103. In this state, the control unit 101 performs processing as the patent analysis support device 10 of this embodiment according to the program stored in the storage unit 103.

[0046] In the above description, a program describing the processing in the patent analysis support device 10 is provided using a CD-ROM or DVD-ROM as the recording medium. However, the system is not limited to this, and depending on the configuration of the computer system, the capacity of the program to be provided, a program provided via a transmission medium such as the Internet may be used.

[0047] The program of this embodiment causes a computer system to perform, for example, the steps of: receiving input from a user for a central phrase to be used as the focus of the analysis of a patent document; extracting supplementary phrases related to the central phrase from each of the multiple patent documents to be analyzed; and outputting the central phrase and supplementary phrases.

[0048] The extraction unit 14 shown in Figure 1 is realized by the execution of a computer program stored in the storage unit 103 shown in Figure 6 by the control unit 101 shown in Figure 6. The storage unit 103 shown in Figure 6 is also used to realize the extraction unit 14. The receiving unit 11 shown in Figure 1 is realized by at least one of the input unit 102 and the communication unit 105 shown in Figure 6. The information storage unit 13 shown in Figure 1 is part of the storage unit 103 shown in Figure 6. The document acquisition unit 12 shown in Figure 1 is realized by the communication unit 105 shown in Figure 6. The display unit 15 shown in Figure 1 is realized by the display unit 104 shown in Figure 6. When the display unit 15 performs processing such as generating display data, the control unit 101 and the storage unit 103 are also used to realize the display unit 15. Furthermore, when the patent analysis support device 10 outputs by transmitting the central text and supplementary text to another device, the output unit that performs the output is realized by the communication unit 105 shown in Figure 6. When the output unit that transmits the main text and supplementary text to other devices generates display data, the control unit 101 and the storage unit 103 are also used to realize this output unit. The patent analysis support device 10 may be implemented by multiple computer systems. Furthermore, for example, the patent analysis support device 10 may be implemented by a cloud computing system.

[0049] Furthermore, when the patent analysis support device 10 transmits the central text and supplementary text to a terminal device, the terminal device is also implemented, for example, by a computer system as illustrated in Figure 6. The information processing device that generates the aforementioned trained model, the model storage unit 142 and the inference unit 143 are also implemented, for example, by a computer system as illustrated in Figure 6.

[0050] As described above, in this embodiment, the patent analysis support device 10 extracts supplementary texts that are connected to the central text specified by the user from the document to be analyzed, and outputs the central text and supplementary texts. Therefore, there are no particular restrictions on the document to be analyzed, and the patent analysis support device 10 can present a wide range of information to the user.

[0051] Embodiment 2. Next, Embodiment 2 will be described. The configuration of the patent analysis support device 10 in this embodiment is the same as in Embodiment 1. The operation of this embodiment is the same as in Embodiment 1, except that the extraction targets of the extraction unit 14 differ in some respects from Embodiment 1. Below, the parts that differ from Embodiment 1 will be mainly described, and explanations that overlap with Embodiment 1 will be omitted.

[0052] In this embodiment, the extraction unit 14 extracts supplementary text connected to the central text from the literature to be analyzed, similar to Embodiment 1. In this embodiment, the extraction unit 14 further extracts supplementary text connected to the extracted supplementary text with respect to at least one of the literature to be analyzed. That is, if the supplementary text connected to the central text extracted from the literature to be analyzed is designated as the first supplementary text, the extraction unit 14 extracts a second supplementary text connected to the first supplementary text. The method for extracting the second supplementary text connected to the first supplementary text is the same as the method for extracting the first supplementary text connected to the central text described in Embodiment 1. The second supplementary text may be extracted for all of the literature to be analyzed, or for some of the literature.

[0053] The display unit 15 displays the central text, the first supplementary text, and the second supplementary text. As in Embodiment 1, the central text, the first supplementary text, and the second supplementary text may be output by being transmitted to another device.

[0054] Figure 7 shows an example of a display screen for the central phrase, the first supplementary phrase, and the second supplementary phrase in this embodiment. In the example shown in Figure 7, the first supplementary phrase connected to the central phrase is extracted from each of the analyzed documents, Document 1 and Document 2, and the second supplementary phrase connected to the first supplementary phrase is extracted. In Figure 7 as well, words enclosed in solid lines are the central phrase, and words enclosed in dashed lines are the supplementary phrase.

[0055] In the example shown in Figure 7, the first supplementary text is connected to the central text, "thermal image," by a line. For example, "camera" and "human perceived temperature," extracted from reference 1, are connected to the central text, "thermal image," by lines. Similarly, the second supplementary text is connected to its corresponding first supplementary text by a line. For example, the second supplementary text, "CCD," which is related to the first supplementary text, "camera," is connected by a line. By displaying not only the relationship between the central text and the first supplementary text, but also the relationship between the second supplementary text and the first supplementary text in a visually easy-to-understand manner, users can more easily grasp the relationships between each term. Furthermore, by displaying not only the first supplementary text but also the second supplementary text, users can consider the potential applications of the technological elements included in their patents from a broader perspective.

[0056] In the example described above, an example of extracting a second supplementary phrase connected to the first supplementary phrase was explained. However, the extraction unit 14 may also extract a third supplementary phrase connected to the second supplementary phrase from the document under analysis. For example, the extraction unit 14 may define the extraction of the first supplementary phrase connected to the central phrase from the document under analysis as the first level of extraction, the extraction of the second supplementary phrase connected to the first supplementary phrase from the document under analysis as the second level of extraction, and the extraction of the third supplementary phrase connected to the second supplementary phrase from the document under analysis as the third level of extraction. The extraction unit 14 may perform extraction at multiple levels, and the number of levels is not limited to two but may be three or more. Furthermore, the user may be able to specify how many levels of extraction to perform.

[0057] Embodiment 3. Figure 8 shows an example of the configuration of the patent analysis support device 10a according to Embodiment 3. The patent analysis support device 10a of this embodiment has a cluster identification unit 16 and a document selection unit 17 added to the patent analysis support device 10 described in Embodiment 1. Hereafter, components having the same functions as in Embodiment 1 are denoted by the same reference numerals as in Embodiment 1, and redundant descriptions are omitted. The following will mainly describe the differences from Embodiment 1.

[0058] In Embodiment 1, the patent analysis support device 10 received input from the user for document identification information to be analyzed. In this embodiment, however, the patent analysis support device 10a selects at least a portion of the documents to be analyzed. For example, in this embodiment, the information storage unit 13 stores patent maps (hereinafter also referred to as maps) that show the relationships between patents, such as network diagrams showing citation relationships between patent documents and IP (Intellectual Property) landscape maps. A network diagram represents the chain of relationships between patent documents in a network-like manner. The cluster identification unit 16 identifies clusters composed of documents in similar technical fields based on the maps stored in the information storage unit 13.

[0059] The document selection unit 17 selects at least one of several documents to be analyzed. Specifically, the document selection unit 17 selects one or more patent documents from each of different technical fields as documents to be analyzed. For example, the document selection unit 17 selects documents to be analyzed from each of several clusters. That is, it selects patent documents belonging to different clusters as documents to be analyzed. The document acquisition unit 12 acquires the documents to be analyzed selected by the document selection unit 17. The operation of the extraction unit 14 and the display unit 15 is the same as in Embodiment 1. Here, an example is described in which the information storage unit 13 stores a map, but it is not limited to this, and the cluster identification unit 16 may acquire a map by receiving a map from another device and identify clusters consisting of documents in similar technical fields based on the map.

[0060] Figure 9 shows an example of a network diagram of this embodiment. The network diagram is a diagram in which, for example, references are represented as nodes, and the nodes are connected by edges that indicate citation relationships. In Figure 9, nodes are shown as rectangles and edges are shown as arrows.

[0061] The network diagram is generated, for example, using citation information that shows the citation relationships between patent applications. A citation relationship refers to the relationship between two patent applications in which the patent publication of one patent application is cited in the examination of the other patent application. More specifically, a citation relationship is defined as, with respect to a certain document X, if document Y is cited in the examination of the patent application relating to document X, then document Y is said to be a cited document of document X. Similarly, if document X is cited in the examination of the patent application relating to document Z, then document Z is said to be a cited document of document X. Furthermore, citation relationships also include cited documents of cited documents, cited documents of cited documents, cited documents of cited documents, and cited documents of cited documents. Note that the examination referred to here may include not only the substantive examination of national patent offices, but also the ISR (International Search Report) for PCT applications and the EESR (Extended European Search Report) for EPC (European Patent Convention) applications.

[0062] For example, suppose we have two nodes, one as the first node and the other as the second node, and the first and second nodes are connected by an edge. If this edge is an arrow pointing from the first node to the second node, then this edge indicates that the patent publication of the patent application corresponding to the document indicated by the first node was cited in the examination of the patent application corresponding to the document indicated by the second node. In other words, the edge indicates that there is a citation relationship between the first and second nodes, where the document indicated by the first node is cited in the patent application of the document indicated by the second node. In a network diagram, the direction of the arrow, which is the edge, can be used to distinguish between a citation and a cited document.

[0063] Specifically, a network diagram is generated by, for example, using a certain document as a starting point, and creating nodes representing documents that have a citation relationship with the starting document. Then, the generated nodes are connected by edges to the node representing the patent information that was the starting point. The process of generating nodes based on citation information and connecting nodes with citation relationships by edges is repeated, thereby generating the network diagram. The method of generating the network diagram is not limited to this example.

[0064] In the example shown in Figure 9, cluster 31 and cluster 32 are identified as clusters. The cluster identification unit 16 may, for example, determine an index and identify clusters based on that index. That is, the cluster identification unit 16 determines the nodes belonging to each cluster by identifying the location of the clusters in the network diagram based on the index. The index that the cluster identification unit 16 determines is, for example, a centrality index. The centrality index is an index that quantifies the degree to which each node plays a central role in the network. The cluster identification unit 16 calculates a centrality index such as degree centrality, neighbor centrality, eigenvector centrality, or PageRank. The cluster identification unit 16 outputs the calculated centrality index for each document.

[0065] Degree centrality is an index that represents the degree, which is the number of edges each node has, and nodes with more edges are given a higher evaluation. Degree centrality can also be said to be an index that highly values ​​nodes that have relationships with many other nodes. By calculating degree centrality, the cluster identification unit 16 can analyze the frequency of citations or cited by each node.

[0066] Proximity centrality is an index calculated from the average distance between a given node and all other nodes, with a shorter average distance resulting in a higher rating. Proximity centrality can also be considered an indicator of the ease of access to the entire network. By calculating proximity centrality, the cluster identification unit 16 can analyze whether each node corresponds to a central or representative patent in the network.

[0067] Eigenvector centrality is an index that highly values ​​nodes connected to central nodes, such as nodes with high degree centrality. By determining eigenvector centrality, the cluster identification unit 16 can analyze whether each node corresponds to a promising patent with many connections to central patents.

[0068] PageRank is an extension of eigenvector centrality and is an index that emphasizes the directionality of nodes. By calculating the PageRank, the cluster identification unit 16 can analyze whether each node corresponds to a powerful patent that is a source of citations. Note that the cluster identification unit 16 may also use other centrality indicators in the analysis of the network diagram.

[0069] The method for identifying clusters is not limited to the examples described above. Furthermore, while this example describes how the patent analysis support device 10a identifies clusters, it is not limited to this; the results of cluster identification may be stored in the information storage unit 13 along with the map. Alternatively, the patent analysis support device 10a may receive the results of cluster identification along with the map from another device. In these cases, the patent analysis support device 10a does not need to include a cluster identification unit 16.

[0070] When the network diagram shown in Figure 9 is used, the document selection unit 17 selects one or more documents to be analyzed from each of cluster 31 and cluster 32. Different clusters are likely to belong to different technical fields, and by analyzing documents belonging to different clusters, it is possible to analyze documents from different technical fields. It is not easy for users to compare the contents of documents from different technical fields, but by analyzing documents from different technical fields in this way, it becomes possible to compare supplementary text, making it easier for users to consider the transfer of technical elements between different technical fields.

[0071] The selection of literature from each cluster can be done in any way, but for example, the literature selection unit 17 may select literature that has a citation relationship with each other near the cluster boundary as the subject of analysis, as shown in pair 33 in Figure 9. Alternatively, for example, the literature selection unit 17 may select literature that has a citation relationship with each other and has different IPCs at the top. For example, among the combinations of literature with a citation relationship, the literature selection unit 17 may select the one where the difference in the IPC symbols is at the top when comparing the IPCs of the citation-related literature. Alternatively, the literature selection unit 17 may define the position from the top of the IPC symbols and select combinations of literature with a citation relationship that have different IPCs from the top of the list to that position, using any method, as the subject of analysis.

[0072] Furthermore, the document selection unit 17 may select documents with low similarity in the title of the invention from among the cited documents as documents belonging to a different technical field for analysis. Alternatively, the document selection unit 17 may select documents with low correlation when comparing the similarity of claims using a large-scale language model as documents in a different technical field. For example, the document selection unit 17 may use a large-scale language model to determine the similarity of claims from among the cited documents and select the document with the lowest similarity as the document for analysis. Or, the document selection unit 17 may select documents for analysis from among documents with similarity below a certain value. Furthermore, the document selection unit 17 may collect information other than the document itself based on the bibliographic information of the document and determine whether the technical fields are different based on the collected information. For example, the document selection unit 17 may collect information on the industry of the applicant corresponding to the document, determine whether the technical fields are different based on the similarity of the industry, and select documents in different technical fields as the document for analysis. Information other than literature used to determine whether the technical fields are different may be collected by the literature selection unit 17 via the internet or other means, or it may be acquired in advance and stored in the information storage unit 13.

[0073] Furthermore, one of the documents to be analyzed may be specified by the user, as in Embodiment 1. In this case, the document selection unit 17 selects other documents to be analyzed from a different cluster than the specified document. In this case as well, among the documents that have a citation relationship with the specified document, documents with a higher IPC and different rankings may be selected as the documents to be analyzed.

[0074] The literature selection unit 17 outputs literature identification information of the selected literature to be analyzed to the literature acquisition unit 12. The literature acquisition unit 12 acquires the literature using the literature identification information, similar to Embodiment 1, and stores the acquired literature in the information storage unit 13. The operation of this embodiment, other than that described above, is the same as in Embodiment 1.

[0075] Figure 10 shows an example of a display screen for the central and supplementary texts in this embodiment. In Figure 10, when clusters A and B exist on a map such as a network diagram, documents A1-A5 belonging to cluster A and documents B1-B5 belonging to cluster B are selected as the subjects of analysis. In Figure 10, words enclosed in solid lines are central texts, and words enclosed in dashed lines are supplementary texts. The left side of Figure 10 shows the supplementary and central texts extracted from documents A1-A5, and the right side of Figure 10 shows the supplementary and central texts extracted from documents B1-B5. In Figure 10, five documents are selected from each cluster, but the number of documents selected is not limited to this example. Also, the number of documents selected may differ between clusters.

[0076] Furthermore, while Figure 10 shows a diagram illustrating a single connection by grouping supplementary text extracted from multiple documents belonging to the same cluster, it is not limited to this. For example, the central text and its corresponding supplementary text may be displayed for each document. Also, as shown in Figure 10, when supplementary text extracted from multiple documents belonging to the same cluster is displayed in a diagram illustrating a single connection, the supplementary text may be displayed in different ways for each document. For example, the supplementary text may be displayed in different colors for each document, or the shape of the figure surrounding the supplementary text may differ for each document. The display methods for supplementary text for each document are not limited to these examples.

[0077] Figure 9 shows an example where a network diagram is used as the map, but the map could also be a heat map that shows the relationships between technologies obtained by analyzing literature using some indicator, such as an IP landscape map.

[0078] Figure 11 shows an example of a heat map of this embodiment. In the example shown in Figure 11, the number of documents is indicated by color. In the example shown in Figure 11, there are islands 41 to 44 in the heat map that have a larger value than the others. The cluster identification unit 16 may identify these islands 41 to 44 as a cluster. In this case, for example, the document selection unit 17 selects documents to be analyzed from different islands 41 to 44 in order to select documents from different technical fields to be analyzed. In this case as well, the document selection unit 17 may select documents from different technical fields based on IPC, similarity of invention titles, similarity of claims, information other than documents, etc. The document selection unit 17 may also select documents close to the boundaries of the islands as documents to be analyzed.

[0079] Furthermore, the literature selection unit 17 may select literature from different technical fields using both the network diagram illustrated in Figure 9 and the heatmap illustrated in Figure 11. For example, the literature selection unit 17 may select literature belonging to different islands 41-44 in the heatmap from among the literature with citation relationships in the network diagram as literature to be analyzed in different technical fields.

[0080] The maps shown in Figures 9 and 11 are illustrative, and the maps used by the patent analysis support device 10a are not limited to these. Furthermore, the document selection method described above is illustrative, and the document selection method using the cluster identification results is not limited to the examples described above. Moreover, the method for determining whether the technical fields are different is not limited to the example using the maps described above. For example, the document selection unit 17 may determine whether the technical fields are different based on the IPC of each document, the similarity of the invention title, the similarity of the claims, and other information besides the document, and select documents from different technical fields as documents to be analyzed. In this case, the patent analysis support device 10a does not need to include the cluster identification unit 16.

[0081] The patent analysis support device 10a of this embodiment is implemented, similar to the patent analysis support device 10 of Embodiment 1, by, for example, the computer system illustrated in Figure 6. The extraction unit 14 shown in Figure 8 is implemented by executing a computer program stored in the storage unit 103 shown in Figure 6 by the control unit 101 shown in Figure 6. The storage unit 103 shown in Figure 6 is also used to implement the extraction unit 14.

[0082] Furthermore, the patent analysis support device 10a of this embodiment may perform extraction of multiple layers, as described in Embodiment 2.

[0083] As described above, in this embodiment, the patent analysis support device 10a is configured to select documents from different technical fields as analysis targets. This makes it easier for the user to understand the contents of documents from different technical fields. Furthermore, if the user specifies a document in which they are considering the application of a technical element as one of the analysis targets, it becomes easier to consider the application of that technical element to a different technical field.

[0084] Embodiment 4. Next, Embodiment 4 will be described. The configuration of the patent analysis support device 10a in this embodiment is the same as in Embodiment 3. The operation of this embodiment is the same as in Embodiment 3, except that the document selection unit 17 selects documents that are highly relevant (highly related to each other) instead of documents in different technical fields. Below, we will mainly describe the parts that differ from Embodiments 1 and 3, and omit explanations that overlap with Embodiments 1 and 3.

[0085] In Embodiment 3, documents from different technical fields were analyzed. However, depending on the purpose of the study, the user may want to compare documents from highly relevant technical fields. In this embodiment, for example, the document selection unit 17 selects documents to be analyzed from documents within the same cluster. This allows the user to compare the supplementary text of highly relevant documents, making it easier to compare the content of the documents. In this case as well, one of the documents to be analyzed may be specified by the user, as in Embodiment 1. Furthermore, highly relevant documents are not limited to documents belonging to the same cluster. As described in Embodiment 3, they may be determined based on the IPC of each document, the similarity of the invention title, the similarity of the claims, the presence of the same keywords, or other information besides the document. For example, based on at least one of the similarity of the invention title, the similarity of the claims, or other information besides the document, documents may be deemed highly relevant if they have a high similarity, or if the applicants' industries are the same or similar. Furthermore, the document selection unit 17 may determine that a document is highly relevant not only if it belongs to the same cluster, but also if it belongs to different clusters, based on at least one of the following: similarity of the invention title, similarity of the claims, or information other than the document itself, such as if the similarity is high or if the applicant's industry is the same or similar. Alternatively, the user may specify a reference document and then specify conditions that are highly relevant to the reference document as search conditions, and the document selection unit 17 may select highly relevant documents. The document selection unit 17 may also use a generating AI, in which case the search conditions may be specified as a prompt.

[0086] Furthermore, as described in Embodiment 3, the user may be able to select literature from a different technical field, or as described in this embodiment, they may be able to select literature from a highly relevant technical field. Also, when selecting literature from a different technical field, the user may be able to select literature from fields that are as far apart as possible, or literature from fields that are close together. Literature from fields that are as far apart as possible may be, for example, literature with different IPC rankings, or literature with different IPC rankings but matching top rankings, but are not limited to these. Literature from fields that are as far apart as possible may also be, for example, literature with distant cluster centers. Thus, whether a field is far apart or related may be determined based on the distance between the cluster centers.

[0087] Furthermore, in this embodiment as well, the patent analysis support device 10a of this embodiment may perform extraction of multiple layers, as described in Embodiment 2.

[0088] As described above, in this embodiment, the patent analysis support device 10a is configured to select highly relevant documents as analysis targets. This makes it easier for the user to understand the content of highly relevant documents. Furthermore, if the user specifies a document in which they are considering the application of a technical element as one of the analysis targets, it becomes easier to consider the application of that technical element. In addition, when considering the application of a technical element to a different technical field, if highly relevant documents from among documents belonging to different clusters are selected as analysis targets, the user can obtain information on highly relevant documents even among documents in different technical fields, thus efficiently considering the application.

[0089] Embodiment 5. Next, Embodiment 5 will be described. The configuration of the patent analysis support device 10 in this embodiment is the same as in Embodiment 1. The operation of this embodiment is the same as in Embodiment 1, except that the patent analysis support device 10 extracts multiple key phrases from the document to be analyzed. Below, we will mainly describe the parts that differ from Embodiment 1, and omit explanations that overlap with Embodiment 1.

[0090] In this embodiment, the reception unit 11 receives input of multiple central phrases. The extraction unit 14 extracts related supplementary phrases from the document under analysis for each of the multiple central phrases. The display unit 15 displays the multiple central phrases and their corresponding supplementary phrases. Similar to Embodiment 1, the patent analysis support device 10 may output the multiple central phrases and their corresponding supplementary phrases by inputting them to another device.

[0091] Figure 12 shows an example of the display screen for the central and supplementary texts in this embodiment. In Figure 12, the extraction results for Document 1 are displayed at the top, and the extraction results for Document 2 are displayed at the bottom. In Figure 12, words enclosed in solid lines are central texts, and words enclosed in dashed lines are supplementary texts. In Figure 12, "thermal image" and "airflow" are designated as central texts. The number of central texts is not limited to the example in Figure 12 and may be multiple.

[0092] In the example shown in Figure 12, each central phrase is displayed with its associated supporting phrases connected by lines. This method of connecting central phrases with their corresponding supporting phrases makes it easier for users to understand the relationship between them. Furthermore, displaying the supporting phrases of two documents being analyzed on a single screen makes it easier for users to compare the content of the documents. Note that the method of displaying central and supporting phrases is not limited to the example shown in Figure 12. For example, supporting phrases may be displayed for each document rather than separately for each central phrase. For instance, both "thermal image" and "airflow" may be displayed within a rectangle representing the central phrase, and the supporting phrases corresponding to "thermal image" and "airflow" may be displayed connected by lines from this rectangle. The method of displaying central and supporting phrases is not limited to these examples.

[0093] In this embodiment, the patent analysis support device 10 accepts multiple core phrases and extracts supplementary phrases for each of the multiple core phrases, allowing the user to compare documents from a more multifaceted perspective. Furthermore, when the user designates a reference document corresponding to their own patent as the target of analysis, they can examine from multiple perspectives whether the technical elements contained in the reference document can be applied to different technical fields.

[0094] Furthermore, in this embodiment as well, the patent analysis support device 10 of this embodiment may perform extraction of multiple levels, as described in Embodiment 2. Also, the patent analysis support device 10a described in Embodiment 3 or Embodiment 4 may accept input of multiple central phrases and extract supplementary phrases for each of the multiple central phrases, as described in this embodiment.

[0095] The configurations shown in the above embodiments are merely examples, and it is possible to combine them with other known technologies, combine different embodiments, and omit or modify parts of the configuration without departing from the gist of the invention.

[0096] The various aspects of this disclosure are summarized below as an appendix.

[0097] (Note 1) A reception unit that receives input from users for the central phrases to be used as the basis for the analysis of patent documents, An extraction unit that extracts supplementary phrases related to the central phrase from each of the multiple patent documents to be analyzed, An output unit that outputs the aforementioned central text and the aforementioned supplementary text, A patent analysis support device characterized by comprising the following features. (Note 2) A document selection unit that selects at least one of the multiple patent documents to be analyzed. The patent analysis support device according to Appendix 1, characterized by comprising the above. (Note 3) The patent analysis support device according to Appendix 2, characterized in that the document selection unit selects one or more patent documents from each of different technical fields as the patent documents to be analyzed. (Note 4) The document selection unit selects patent documents belonging to different clusters in the patent map as the patent documents to be analyzed. The patent analysis support device according to Appendix 3, characterized by comprising the above. (Note 5) The patent analysis support device according to Appendix 4, characterized in that the patent map includes a network diagram showing the citation relationships between patent documents. (Note 6) The patent analysis support device according to Appendix 5, characterized in that the document selection unit uses the network diagram to select patent documents belonging to different clusters from among the patent documents in a citation relationship as the patent documents to be analyzed. (Note 7) The patent analysis support device according to any one of the appendices 2 to 6, characterized in that the document selection unit selects the patent documents to be analyzed such that the patent documents to be analyzed include patent documents that are highly related to each other. (Note 8) The reception unit receives from the user the designation of one or more reference documents from the patent documents to be analyzed. The patent analysis support device according to any one of the appendices 1 to 7, characterized in that the extraction unit extracts supplementary texts related to the central phrase from each of the plurality of patent documents to be analyzed, including the reference document specified by the user. (Note 9) The patent analysis support device according to any one of the appendices 1 to 8, characterized in that the extraction unit extracts a second ancillary phrase which is connected to the first ancillary phrase which is the ancillary phrase. (Note 10) The reception unit receives input of multiple central phrases from the user. The patent analysis support device according to any one of the appendices 1 to 9, characterized in that the extraction unit extracts the supplementary phrases that are connected to each of the plurality of central phrases. (Note 11) In a computer system, A step in which the user inputs the central phrase to be used as the basis for the analysis of the patent document, The steps include: extracting supplementary phrases related to the central phrase from each of the multiple patent documents to be analyzed; The steps include outputting the main statement and the supplementary statement, A program characterized by causing the execution of a specific action. (Note 12) A patent analysis support method performed by a patent analysis support device, A step in which the user inputs the central phrase to be used as the basis for the analysis of the patent document, The steps include: extracting supplementary phrases related to the central phrase from each of the multiple patent documents to be analyzed; The steps include outputting the main statement and the supplementary statement, A patent analysis support method characterized by including the following. [Explanation of symbols]

[0098] 10,10a Patent analysis support device, 11 Reception unit, 12 Document acquisition unit, 13 Information storage unit, 14 Extraction unit, 15,104 Display unit, 16 Cluster identification unit, 17 Document selection unit, 31,32 Cluster, 101 Control unit, 102 Input unit, 103 Storage unit, 105 Communication unit, 106 Audio output unit, 107 System bus, 141 Learning unit, 142 Model storage unit, 143 Inference unit.

Claims

1. A reception unit that receives input from users for the central phrases to be used as the basis for the analysis of patent documents, An extraction unit that extracts supplementary phrases related to the central phrase from each of the multiple patent documents to be analyzed, An output unit that outputs the aforementioned central text and the aforementioned supplementary text, A patent analysis support device characterized by comprising the following features.

2. A document selection unit that selects at least one of the multiple patent documents to be analyzed. The patent analysis support device according to claim 1, characterized by comprising the above.

3. The patent analysis support apparatus according to claim 2, characterized in that the document selection unit selects one or more patent documents from each of different technical fields as the patent documents to be analyzed.

4. The document selection unit selects patent documents belonging to different clusters in the patent map as the patent documents to be analyzed. The patent analysis support device according to claim 3, characterized by comprising the above.

5. The patent analysis support device according to claim 4, characterized in that the patent map includes a network diagram showing the citation relationships between patent documents.

6. The patent analysis support device according to claim 5, characterized in that the document selection unit uses the network diagram to select patent documents belonging to different clusters from among the patent documents in a citation relationship as the patent documents to be analyzed.

7. The patent analysis support apparatus according to claim 2, characterized in that the document selection unit selects the patent documents to be analyzed such that the patent documents to be analyzed include patent documents that are highly related to each other.

8. The reception unit receives from the user the designation of one or more reference documents from the patent documents to be analyzed. The patent analysis support device according to any one of claims 1 to 7, characterized in that the extraction unit extracts supplementary texts related to the central phrase from each of the plurality of patent documents to be analyzed, including the reference document specified by the user.

9. The patent analysis support apparatus according to any one of claims 1 to 7, characterized in that the extraction unit extracts a second ancillary phrase which is connected to the first ancillary phrase which is the ancillary phrase.

10. The reception unit receives input of multiple central phrases from the user. The patent analysis support apparatus according to any one of claims 1 to 7, characterized in that the extraction unit extracts the supplementary phrases that are connected to each of the plurality of central phrases.

11. In a computer system, A step in which the user inputs the central phrase to be used as the basis for the analysis of the patent document, The steps include: extracting supplementary phrases related to the central phrase from each of the multiple patent documents to be analyzed; The steps include outputting the main statement and the supplementary statement, A program characterized by causing the execution of a specific action.

12. A patent analysis support method performed by a patent analysis support device, A step in which the user inputs the central phrase to be used as the basis for the analysis of the patent document, The steps include: extracting supplementary phrases related to the central phrase from each of the multiple patent documents to be analyzed; The steps include outputting the main statement and the supplementary statement, A patent analysis support method characterized by including the following.

Images