Technical literature search system, compound search system, and technical literature search method

Abstract

The system automatically searches for technical documents or compounds that possess the properties desired by the user. [Solution] A technical literature search system comprising an information processing device having an input interface and a processor, and a technical literature database, further comprising a characteristic value database storing information where characteristic values ​​and technical literature are associated, wherein the input interface receives at least one of the maximum and minimum values ​​of the characteristic values ​​as input, the processor retrieves technical literature that satisfies the input characteristic values ​​from the technical literature database, the input interface receives keyword selection or input, and the processor extracts from the technical literature database those that contain the keyword.

Description

Technical Field

[0001] The present disclosure relates to a technical document search system, a compound search system, and a technical document search method.

Background Art

[0002] In recent years, in the field of material development, efforts have been made to efficiently develop promising materials, material compositions, etc. from the perspectives of accelerating development for early market entry of new products, reducing costs and environmental considerations by reducing power consumption and waste, and achieving both labor saving and work-life balance. In the efficiency improvement of material development, materials informatics that analyzes material data using information engineering to search for high-performance materials has attracted international attention.

[0003] However, in order to incorporate materials informatics into material development, it is necessary for the user to prepare data for material search in a rich and analyzable format such as by machine learning. Regarding this problem, a method has been proposed to utilize a vast number of already published technical information such as patents and papers to search for high-performance materials.

[0004] Patent Document 1 discloses a compound search system for finding biologically active compounds, which includes a compound searcher that calculates the feature vector distance between the feature vector of a specified compound recorded in a model table and the feature vector of each compound recorded in a search table, and retrieves similar compounds according to the feature vector distance. Patent Document 1 states that such a search system makes it possible to search for compounds with high similarity in terms of both the number of structural features and physicochemical properties with high accuracy. Patent Document 1 also states that public databases can be used as dedicated compound databases for each field and general compound databases, and that these databases record, for example, ID (or information identifying the compound), name, structural formula (e.g., InChI: International Chemical Identifier), physicochemical properties, biological activity, etc., for multiple compounds.

[0005] Patent Document 2 discloses a numerical search device for searching numerical data related to numerical values ​​contained in a document, comprising a data extraction unit and a database unit. The data extraction unit divides the collected text data of the patent document into multiple morphemes, identifies predetermined feature parts and numerical parts from the divided morphemes, and determines the relationship between the feature parts and numerical parts, but pre-sets conditions to exclude from the determination of the relationship between the feature parts and numerical parts. Patent Document 2 states that such a numerical search device can correctly associate and extract numerical parts and feature parts. [Prior art documents] [Patent Documents]

[0006] [Patent Document 1] Japanese Patent Publication No. 2021-108108 [Patent Document 2] Japanese Patent Publication No. 2020-80087 [Overview of the Initiative] [Problems that the invention aims to solve]

[0007] In the search system described in Patent Document 1, a search module (compound searcher) acquires feature vectors representing various structural or physicochemical properties, and uses these feature vectors to search for compounds. However, when inputting user-specified compounds into the search module, the user had to arbitrarily select the compounds to input from their own knowledge or known technical literature. As a result, the method of selecting compounds to input into the search module depended on the user, and there was a risk that the desired compounds would not be extracted in the output results, which was affected by the user's skills. In addition, the selection of input compounds before compound search took time, resulting in an overall problem of time being required for material search.

[0008] The numerical search device described in Patent Document 2 divides the collected text data of patent documents into multiple morphemes, identifies predetermined feature parts and numerical parts from the divided morphemes, and determines the relationship between the feature parts and numerical parts. However, the method for determining the relationship between predetermined feature parts and numerical parts is excellent for obtaining matching information, but it has the problem that it cannot extract information that satisfies a certain numerical range.

[0009] The purpose of this disclosure is to automatically search for technical documents or compounds that possess the characteristics desired by the user. [Means for solving the problem]

[0010] The technical document search system of this disclosure comprises an information processing device having an input interface and a processor, a technical document database, and further comprises a characteristic value database storing information relating characteristic values ​​to technical documents, wherein the input interface receives at least one of the maximum and minimum values ​​of the characteristic values ​​as input, the processor retrieves technical documents that satisfy the input characteristic values ​​from the technical document database, the input interface receives keyword selection or input, and the processor extracts from the technical document database those that contain the keyword.

[0011] Other forms of this disclosure will be described in the embodiments described below. [Effects of the Invention]

[0012] According to this disclosure, it is possible to automatically search for technical documents or compounds that have the properties desired by the user. [Brief explanation of the drawing]

[0013] [Figure 1A] This is a configuration diagram showing the technical literature search system according to Example 1. [Figure 1B] This is a configuration diagram showing the hardware of the technical document search system according to Example 1. [Figure 2] This table shows an example of a characteristic data table related to Example 1. [Figure 3] This figure shows an example of an input screen according to Example 1. [Figure 4] This figure shows an example of the output screen according to Example 1. [Figure 5] This is a flowchart illustrating the operation of a technical document search using the technical document search system according to Example 1. [Figure 6] This is a diagram showing the compound discovery system according to Example 2. [Figure 7] This table shows an example of a compound data table related to Example 2. [Figure 8]It is a flowchart showing the operation of compound search using the compound search system according to Example 2.

BEST MODE FOR CARRYING OUT THE INVENTION

[0014] Hereinafter, examples of the present disclosure will be described. However, the present disclosure is not limited to the following content and can be arbitrarily modified and implemented without significantly impairing the effects of the present disclosure. The present disclosure can be implemented by combining different embodiments.

EXAMPLE

[0015] <Technical Document Search System> First, the technical document search system in Example 1 of the present disclosure will be described.

[0016] FIG. 1A is a configuration diagram showing the technical document search system according to this example.

[0017] In this figure, the technical document search system 100 is an information processing device 200 which is a computer including an input interface 210, a processor 220, and an output interface 230, a technical document database 310 storing information on technical documents publicly available in the world, and a characteristic value database 320 storing information in which characteristic values of products, materials, etc. are associated with technical documents. In this specification, "information" is assumed to be included in "data". "Data" refers to what can be processed by a computer such as arithmetic operations, stored by a storage device, etc., and includes not only what can be processed by a digital computer but also what can be processed by a quantum computer.

[0018] The input interface 210 is a configuration for inputting characteristics, keywords, etc., into a computer system, and is not particularly limited. Specifically, it can be a keyboard, mouse, touch panel, etc. The information input through the input interface 210 is displayed on a visual display device such as a display or monitor. The input interface 210 may be built into the information processing device 200 or installed outside the information processing device 200.

[0019] The processor 220 is a central component of a computer, a device that processes data and instructions, and often refers to the central processing unit (CPU), but is not particularly limited to it. The processor 220 is a crucial component that greatly affects the performance and speed of the computer, and performance improves as the clock frequency increases. Also, if the processor 220 has multiple cores, it can process multiple tasks simultaneously, improving multitasking performance. Furthermore, the more threads the processor 220 has, the higher its parallel processing capability. The larger the cache memory, the faster the data access speed. Thus, the processing power of the technical literature search system 100 is determined by the processor 220. It is desirable that the processor 220 have a configuration suitable for technical literature search processing.

[0020] The output interface 230 is configured for transmitting data from the information processing device 200 to an external device or system, and is not particularly limited. The information output through the output interface 230 can be displayed on a display or monitor, which is a device for visual display. The output interface 230 may be built into the information processing device 200 or installed outside of the information processing device 200.

[0021] In addition to the input interface 210, processor 220, and output interface 230, the information processing device 200 may also include a storage device for saving data. The storage device has the function of saving information input through the input interface 210, information output through the output interface 230, and so on.

[0022] The technical literature database 310 is a database that stores data on technical literature, including papers, research materials, and documents published in patent gazettes in fields such as science and technology, medicine, and engineering. It is not particularly limited to any database that stores data on technical literature describing compounds. The technical literature database 310 may be a publicly available online service or a database that stores private data. However, in terms of the amount of data, a publicly available online service is more desirable.

[0023] Figure 1B is a configuration diagram showing the hardware of the technical document search system according to this embodiment.

[0024] In this figure, the technical document search system 1000 is implemented using a computer and includes a display 1210, an input device 1220, a CPU 1230 (Central Processing Unit), RAM 1240 (Random Access Memory), ROM 1250 (Read Only Memory), a communication device 1260, a media reader 1270, and an auxiliary storage device 1280. The display 1210 corresponds to the display, monitor, etc. of the output interface 230 in Figure 1A. The input device 1220 corresponds to the input interface 210, etc., in Figure 1A.

[0025] The CPU 1230 is a unit that can be implemented as a so-called processor and performs various calculations. The CPU 1230 performs various processes by executing technical document search processing programs and the like that loaded from the auxiliary storage device 1280 into the RAM 1240. Therefore, the CPU 1230 corresponds to the processor 220 in Figure 1A.

[0026] Here, the technical literature search processing program is, for example, an application program that can be executed on an OS (Operating System) program. The technical literature search processing program may consist of multiple modules, or it may be implemented by multiple independent programs.

[0027] Furthermore, the technical literature search processing program may be installed on the auxiliary storage device 1280 from a portable storage medium via, for example, the media reader 1270. In other words, the technical literature search processing program can be stored on a storage medium. Note that the CPU 1230 may execute processing according to a program other than the technical literature search processing program. In this case, it is desirable to store this program in the auxiliary storage device 1280.

[0028] The RAM 1240 is memory that stores programs such as the technical literature search processing program executed by the CPU 1230, as well as data necessary for program execution. The ROM 1250 is memory that stores programs necessary for starting the technical literature search system 1000, the OS, etc. The communication device 1260 is capable of sending and receiving data with external mobile terminals, etc., via a communication channel. The communication channel can be a network such as a LAN (Local Area Network) or the Internet, and can be wired or wireless.

[0029] The media reader 1270 is a device that reads information from portable storage media such as flash memory and CD-ROMs.

[0030] The auxiliary storage device 1280 can be implemented as, for example, an HDD (Hard Disk Drive) and is a device that stores data and programs for executing various processes. Alternatively, the auxiliary storage device 1280 may be implemented as an SSD (Solid State Drive) using flash memory or the like.

[0031] The RAM 1240, ROM 1250, and auxiliary storage device 1280 correspond to the technical literature database 310 and characteristic value database 320 shown in Figure 1A. The auxiliary storage device 1280 stores the technical literature search processing program, various information, and data.

[0032] Figure 2 is a table showing an example of a characteristic data table related to this embodiment.

[0033] This figure shows a characteristic data table 250, which is part of the characteristic value database 320 that constitutes the technical literature search system 100 in Figure 1A. It is a table of characteristics described in technical literature and their maximum and minimum values.

[0034] As illustrated in Figure 2, the maximum and minimum values ​​for pre-defined characteristic values ​​such as glass transition temperature, dielectric loss tangent, and melt flow rate are extracted and organized. The maximum and minimum values ​​for characteristic values ​​are extracted using natural language processing. For example, for characteristic values ​​that can be identified by their units, such as melt flow rate, all numerical values ​​with corresponding units are extracted, the units are standardized, and then the maximum and minimum values ​​are extracted from the set of characteristic values. On the other hand, for values ​​where it is not possible to determine whether it is glass transition temperature or melting point from the units, the determination is made from the context, and the characteristic and numerical value are associated. Similarly, for unitless characteristics such as dielectric loss tangent, the determination is made from the context, and the characteristic and numerical value are associated. If the numerical value for a pre-defined characteristic is not listed in technical literature, it is recorded as dataless information.

[0035] In this way, the same natural language processing is applied to all the data of the technical documents stored in the technical document database 310 in Figure 1A, and the characteristic data table 250 in Figure 2 is created.

[0036] Figure 3 shows an example of an input screen according to this embodiment.

[0037] This figure shows the input screen 211 of the input interface 210 that constitutes the technical literature search system 100 in Figure 1A.

[0038] The input screen 211 displays a characteristics input area 212 and a keyword input area 213.

[0039] The user can select a characteristic to search for from the characteristic column in the characteristic input area 212, based on pre-configured characteristics. In addition to selecting a characteristic, the user can also select either the maximum or minimum value of the characteristic. Furthermore, the user can optionally specify a range in the condition column where the maximum or minimum characteristic value exists. This range condition can be based on the maximum value only, the minimum value only, or both the maximum and minimum values.

[0040] Furthermore, users can further refine their search by selecting or entering keywords in the keyword input area 213. For example, they can enter keywords such as products (automobiles, batteries, semiconductors, resins, etc.), materials (thermosetting resins, polymers, etc.), or functions (biodegradability, etc.). The selectable keywords are those for which words with the same meaning are pre-registered in the library, and all words registered in the library corresponding to the keyword are included in the search.

[0041] Furthermore, even if only one of the terms "selection" or "input" is mentioned in the following explanation, it is assumed that the other method may also be used.

[0042] When the user clicks the "Start Search" button on the input screen 211, the processor 220 in Figure 1A refers to the characteristic value database 320 and searches the technical document database 310 for technical documents that satisfy the characteristic value entered on the input screen 211. Furthermore, if a keyword is entered in the keyword input area 213, the processor 220 extracts only the technical documents that contain the keyword from the search results.

[0043] Figure 4 shows an example of the output screen according to this embodiment.

[0044] This figure shows the output screen 231 of the output interface 230 that constitutes the technical literature search system 100 in Figure 1A. Finally, the technical literature extracted by the processor 220 is output by the output interface 230. The technical literature output by the output interface 230 is displayed on the output screen 231. The output screen 231 displays unique information that identifies the technical literature and information that allows the user to understand the content of the technical literature. Here, unique information that identifies the technical literature includes the title of the technical literature, the publication date, the country of the author's residence, and the author's affiliation. Information that allows the user to understand the content of the technical literature includes the abstract, etc.

[0045] The display method of the output screen 231 is not particularly limited, and the title of the technical document, abstract, country of residence of the author of the technical document, and affiliation of the author of the technical document can be classified and displayed depending on the processing method of the processor 220.

[0046] Figure 5 is a flowchart illustrating the operation of the technical literature search according to this embodiment.

[0047] In this diagram, the user inputs the desired properties and range of characteristic values ​​for the compound they wish to search for into the input interface 210 in Figure 1A (step S510). Furthermore, the user selects or inputs keywords (free words) to narrow down the technical literature in the input interface 210 (step S520). In other words, the input interface 210 accepts the above inputs.

[0048] The processor 220 extracts technical documents (including patent documents) from publicly available technical documents that satisfy some or all of the user's required characteristics, filters them to include only those with keywords in their title or abstract, and outputs the technical documents requested by the user via the output interface 230 (step S530).

[0049] Here, the method for extracting technical documents is not limited to technical documents in which keywords are included in the title or abstract. If a technical document is structured into chapters or sections, it is possible to extract technical documents in which keywords are included within those chapters or sections. For example, if the technical document is a patent document, blocks of text such as claims and embodiments correspond to chapters or sections. Keywords may also include the country of application, applicant, and post-application status (progress information) listed in the patent document.

[0050] In this way, users can search for the technical literature they need through a simple process. [Examples]

[0051] This embodiment relates to a compound search system that includes components of the technical literature search system according to Example 1. In the following description, parts common to Example 1 will be omitted from the explanation.

[0052] <Compound Discovery System> Figure 6 is a diagram showing the configuration of the compound discovery system according to this embodiment.

[0053] In this figure, the compound search system 600 includes a compound-characteristic database 322. The compound-characteristic database 322 includes a characteristic database 320 and a compound database 321. The characteristic database 320 is the same as that of Example 1. The compound database 321 stores information that associates compounds with technical literature. The characteristic database 320 and the compound database 321 are configured to associate characteristic values, including physical properties, with compounds by linking them through technical literature.

[0054] The input interface 210 is configured to allow input of characteristics and keywords as shown in Figure 1A, as well as unique information linked to technical literature, fields of use, and applications. Furthermore, the input interface 210 is configured to allow input of keywords related to the compound's characteristics, fields of use, or applications.

[0055] The output interface 230 is configured to output not only the technical documents shown in Figure 1A, but also the classification of compounds found in those technical documents.

[0056] In addition to the technical literature search shown in Figure 1A, the processor 220 is configured to perform processes such as compound searching, linking compound classification keywords with compounds, and systematic classification of compounds.

[0057] By utilizing the compound-characteristic database 322, it is possible to search not only for technical documents but also for compounds that possess characteristic values ​​specified by the user, as described in those technical documents.

[0058] The following are some examples of its physical properties.

[0059] Basic physical properties include molecular weight, degree of polymerization, density, viscosity, and loss factor. Biochemical properties include IC50, pesticide usage, and cytotoxicity. Properties related to resin molding include melt flow rate, strength-related properties such as Charpy impact strength, Izod impact strength, Vickers impact hardness, Rockwell C impact hardness, tensile modulus, flexural modulus, yield strength, tensile strength, flexural strength, fracture strength, tensile elongation at fracture, compressive strength, shear strength, adhesive strength, peel strength, sliding resistance, and storage modulus. Properties related to electromagnetics and batteries include dielectric loss tangent (tanδ), relative permittivity, discharge capacity, internal resistance, Li-ion conductivity, polymer electrical conductivity, cycle count, particle size / grain size, capacity retention rate, specific surface area, carbon content, CTI (Comparative Tracking Index), and solar conversion efficiency. Factors related to light include transmission loss, light transmittance, refractive index, absorption wavelength, optical density (OD value), reflectance, brightness, contrast, chromaticity coordinate X, and chromaticity coordinate Y. Factors related to heat include heat resistance temperature (thermal stability), glass transition temperature, melting point, boiling point, thermal expansion coefficient, thermal conductivity, flash point, critical oxygen concentration, curing time, curing temperature, curing shrinkage, and softening temperature. Factors related to semiconductors include exposure wavelength, exposure amount, gas reactivity (Å / cycle), etching rate, deposition rate, vapor pressure, and polishing rate. Factors related to quantum mechanics include HOMO, LUMO, PL (photoluminescence) wavelength, and PL (photoluminescence) quantum yield. Factors related to water include contact angle, moisture absorption / water absorption rate, HLB (hydrodrophilic-lipophilic balance), critical micelle concentration (CMC), and SP (solubility parameter). Factors related to gases include water vapor transmittance and oxygen transmittance. Factors related to synthesis and catalysis include reaction yield, activation energy, adsorption amount, adsorption temperature, adsorption time, purification rate, crystallization time, and reaction selectivity.

[0060] Here, we will explain a specific example of adjusting the properties of a compound using the compound-property value database 322.

[0061] For example, if the challenge is to suppress the biodegradability of a resin material that uses polylactic acid as its base material by mixing some kind of additive with polylactic acid, then we will search for an appropriate additive (compound, etc.) using biodegradability as a keyword.

[0062] Furthermore, when the challenge is to improve the heat resistance of a resin composition that is permitted to contain multiple types of resins, we will search for an appropriate resin (compound, etc.) using heat resistance as a keyword.

[0063] Figure 7 is a table showing an example of a compound data table related to this embodiment.

[0064] This figure shows the compound data table 750, which is part of the compound database 321 that constitutes the compound discovery system 600 in Figure 6.

[0065] Compound Data Table 750 is a table of compounds listed in technical literature. It organizes compound names extracted from specified technical literature using natural language processing, along with their corresponding InChI (International Chemical Identifier) ​​formulas. The InChI formula is a unique ID for each compound. Instead of the InChI formula, other information that can identify a compound, such as InChIKey or SMILES, can also be used. The sources of information for Compound Data Table 750 are not particularly limited.

[0066] The compound search system in this embodiment accesses public databases (including those made available to the public via telecommunication lines (such as the Internet)) based on such compound-specific information, extracts corresponding technical literature, and creates a compound data table 750. Examples of public databases that can be used to search for compounds based on such compound-specific information include PubChem, ChemSpider, NIST Chemistry WebBook, and Chemical Identifier Resolver.

[0067] By using data obtained from publicly available databases of compounds, identical or similar compounds can be extracted. This allows for the extraction of identical or similar compounds in terms of both the number of structural features and physicochemical properties.

[0068] Furthermore, the compound database 321 stores information such as the fields of use (automotive, pharmaceutical, biotechnology, etc.) and applications (engines, inhibitors, biodegradable polymers, etc.) for each compound extracted from technical literature associated with the compound using natural language processing. Therefore, compounds extracted by the processor 220 from the compound database 321 can be classified according to the physical properties, fields of use, applications, etc., entered into the input interface 210. This classification is performed by the processor 220.

[0069] There are several methods for natural language processing to filter technical literature by keywords, and these are not particularly limited. For example, rule-based natural language processing techniques can be utilized. These include tokenization techniques that divide sentences into words and phrases, morphological analysis techniques that identify the base form and part of speech of words, which are effective in languages ​​where word boundaries are not clear, such as Japanese, syntactic analysis techniques that analyze the structure of sentences and clarify the relationships between words, semantic analysis techniques that analyze the meaning of words and sentences considering the context, information extraction techniques that automatically extract specific information from literature, such as compound names and physical properties, and text classification techniques that classify documents into predefined categories.

[0070] Natural language processing techniques for filtering technical literature by keywords can also utilize generative AI. For example, GPT (Generative Pre-trained Transformer), which has advanced natural language generation capabilities by pre-training on large amounts of text data and fine-tuning it for specific tasks, and BERT (Bidirectional Encoder Representations from Transformers), which enables bidirectional contextual understanding and is used in text comprehension, classification, and question answering systems.

[0071] By utilizing the natural language processing techniques described above, the processor 220 can extract compounds corresponding to keywords from a limited area within a specified technical document.

[0072] In technical literature, compounds belonging to similar families are often described in close proximity. Therefore, compounds described in close proximity within a text can be classified as belonging to the same family. For example, compounds described in each paragraph of a document can be extracted, and these extracted compounds can be classified into multiple families.

[0073] In technical literature, when describing compounds with defined uses, such as reaction initiators and additives, they are often described according to their specific use. Compounds described near a particular use within the technical literature can be classified into use-related categories. For example, compounds in paragraphs containing keywords can be extracted, and these extracted compounds can be classified into categories related to keywords such as use. This method is not limited to use; classification can be performed using any keyword.

[0074] In technical literature, when compound names share common elements, such as carboxylic acids or amides, they are often described according to the group of compounds sharing these common elements. Therefore, when compounds with common names or substituents appear in technical literature, they can be classified by their names.

[0075] The method for classifying compounds systematically is pre-programmed as a processing method for the processor 220, but a method that can be specified by the user at the input interface 210 may be adopted, or a method that is processed without going through the input interface 210 may be adopted. Both of these methods may be selectable. The processing method of the processor 220 is not limited to adopting one of the above methods or the like.

[0076] By limiting the sections of the compound description in this way, it becomes possible to search for compounds with a higher correlation in terms of physical properties, fields of use, and applications.

[0077] The hardware configuration of the compound discovery system in this embodiment is the same as that shown in Figure 1B.

[0078] Figure 8 is a flowchart showing the compound discovery operation according to this embodiment.

[0079] In this figure, the user inputs unique information linked to technical documents (including patent documents) into the input interface 210 in Figure 6 (step S810). The user also selects or inputs keywords (free word) related to the physical properties, field of use, or application of the compound into the input interface 210 (step S820). In other words, the input interface 210 accepts the above inputs.

[0080] The processor 220 then extracts compounds corresponding to keywords from the technical documents specified by the user. The extracted compounds are output by the output interface 230 (step S830). Here, the specified technical documents refer to those selected by the user from among the multiple technical documents extracted by the operations in steps S810 and S820.

[0081] This disclosure includes the following programs and databases.

[0082] The program is a program for causing a computer to perform the following steps, wherein the computer comprises at least an input interface and a processor, and further comprises a technical literature database and a characteristic value database storing information relating characteristic values ​​to technical literature, or a configuration connected to the technical literature database and the characteristic value database, wherein the input interface performs the step of receiving at least one of the maximum and minimum values ​​of the characteristic value as input, the processor performs the step of obtaining from the technical literature database technical literature that satisfies the input characteristic value, the input interface performs the step of selecting or receiving keywords as input, and the processor performs the step of extracting from the technical literature database that the keywords are contained in the technical literature.

[0083] The characteristic value database is connected to an information processing device equipped with an input interface and a processor, and stores information that associates characteristic values ​​with technical documents.

[0084] A compound database stores information that links compounds with technical literature.

[0085] The effects of this disclosure are described below.

[0086] According to this disclosure, a characteristic value database can be used, which stores information linking characteristic values ​​of compounds, etc., with technical literature. This allows for the automatic searching of technical literature describing compounds, etc., that possess the characteristics desired by the user.

[0087] Furthermore, according to this disclosure, it is possible to use a characteristic value database as well as a compound database that stores information linking compounds and technical literature. This allows users to automatically search for compounds with desired characteristics and the technical literature describing those compounds.

[0088] In this specification, "characteristics" refers to the properties of materials such as products and compounds. Furthermore, "characteristic value" refers to a value that allows for quantitative searching, determination, etc., of those characteristics. Therefore, the maximum, minimum, and average values ​​of characteristic values ​​can serve as indicators representative of the characteristic value. [Explanation of symbols]

[0089] 100, 1000: Technical literature search system, 200: Information processing device, 210: Input interface, 211: Input screen, 212: Characteristic input area, 213: Keyword input area, 220: Processor, 230: Output interface, 231: Output screen, 310: Technical literature database, 320: Characteristic value database, 321: Compound database, 600: Compound search system.

Claims

1. An information processing device comprising an input interface and a processor, It includes a technical literature database, Furthermore, it includes a characteristic value database that stores information linking characteristic values ​​to technical literature. The input interface receives at least one of the maximum and minimum values ​​of the characteristic value as input. The processor retrieves technical documents that satisfy the input characteristic value from the technical document database. The aforementioned input interface allows for the selection or input of keywords. The processor is a technical document search system that extracts technical documents containing the keywords from the technical document database obtained from the technical document database.

2. The technical document search system according to claim 1, wherein the input interface allows specifying at least one of the range of the maximum value and the range of the minimum value.

3. The technical document search system according to claim 1, wherein the processor extracts technical documents in which the keyword is described in a specific part of the technical document.

4. The technical document search system according to claim 3, wherein the aforementioned specific section is at least one of the title and the abstract.

5. The aforementioned technical documents include patent documents, as described in claim 1, for the technical document search system.

6. The technical document search system according to claim 5, wherein the keywords include at least one of the filing country, applicant, and post-filing status from the aforementioned patent document.

7. An information processing device comprising an input interface and a processor, It includes a technical literature database, Furthermore, the system includes a compound-characteristic value database which contains information linking characteristic values ​​with technical literature, and a compound database which contains information linking compounds with the aforementioned technical literature. The input interface is configured to allow selection or input of specific information for identifying the technical document. The aforementioned processor is a compound search system that retrieves from the technical document database those technical documents associated with the specific information, and searches for compounds described in the retrieved technical documents.

8. The aforementioned input interface allows for the selection or input of keywords relating to the compound's properties, field of use, or application. The compound search system according to claim 7, wherein the processor extracts the compound corresponding to the keyword from the compound database.

9. The compound discovery system according to claim 8, wherein the processor classifies the extracted compound according to the characteristics, field of use, or application.

10. The compound discovery system according to claim 7, wherein the processor extracts compounds described in each paragraph of the text from the acquired technical document and classifies the group of compounds consisting of the extracted compounds into multiple systems.

11. The compound discovery system according to claim 8, wherein the processor extracts compounds from paragraphs containing the keyword from the acquired technical documents, and classifies the group of compounds consisting of the extracted compounds into a system related to the keyword.

12. The compound discovery system according to claim 11, wherein the keyword is the characteristic, the field of use, or the application.

13. The compound discovery system according to claim 7, wherein the processor classifies compounds having a common name into a system by the common name if such compounds are described in the acquired technical documents.

14. The compound search system according to claim 7, wherein the processor extracts identical or similar compounds based on the searched compound, using data obtained from a public database of the compound.

15. A method for searching for technical literature using an input interface, a processor, a technical literature database, and a characteristic value database in which information relating characteristic values ​​to technical literature is stored, The input interface includes the step of receiving at least one of the maximum and minimum values ​​of the characteristic value as input, The processor includes the step of obtaining from the technical document database technical documents that satisfy the input characteristic value, The input interface includes a step of receiving keyword selection or input, A method for searching for technical documents, comprising the step of the processor extracting from the technical documents obtained from the technical document database that contain the keyword.

Images