Computer system related to patent search
The computer system enhances patent search systems by generating AI summaries and offering user-friendly features, improving search efficiency and organization management.
Patent Information
- Application Number
- JP2024006492
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-10-11
- Filing Date
- 2024-01-19
- Publication Date
- 2024-11-21
- Estimated Expiration
- 2043-12-04
AI Technical Summary
Existing patent search systems lack innovation and efficiency in providing comprehensive search functionalities and user interaction.
A computer system that utilizes a learning model to generate summary information for patent documents, allowing users to compare text with AI-generated summaries, and includes features like user accounts, organization management, and conditional reporting for enhanced search capabilities.
Enables a novel patent search system with improved user interaction, organization management, and efficient document summarization, facilitating better search and comparison of patent information.
Smart Images

Figure 2024163000000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a computer system for a patent search system. [Background technology]
[0002] 2. Description of the Related Art Patent search systems that perform prior art searches and infringement prevention searches related to patents have been known for some time now (see Non-Patent Document 1). [Prior art documents] [Patent documents]
[0003] [Non-Patent Document 1] “Patent Information Platform J-PlatPat”, [online], [searched on April 28, 2023], Internet<URL:https: / / www.j-platpat.inpit.go.jp / > DISCLOSURE OF THEINVENTION [Problem to be solved by the invention]
[0004] However, there is room for improvement in such patent search systems.
[0005] An object of the present invention is to provide a computer system capable of providing a novel patent search system. [Means for solving the problem]
[0006] The present invention relates to A computer system that causes a computer to execute the following steps: In the input step, conditions for searching document information of a target patent document and date and time information for specifying a frequency of obtaining summary information corresponding to the input conditions are input based on a user's operation; The summary information is information obtained by summarizing a sentence related to the document information using a predetermined learning model, In the acquisition step, the summary information is acquired separately from the text related to the document information; In the transmission step, information indicating that the summary information has been acquired is transmitted to an operation terminal operated by the user according to the date and time information; In the output step, the summary information is output to the operation terminal so that the user can compare the sentence related to the document information with the summary information. It is a computer system. Effect of the Invention
[0007] According to the present invention, it is possible to provide a computer system capable of providing a novel patent search system. [Brief description of the drawings]
[0008] [Figure 1] FIG. 1 is a diagram showing a configuration of a search system in a first embodiment. [Diagram 2] FIG. 2 is a diagram showing a state in which a number search is being performed in the first embodiment. [Diagram 3] FIG. 2 is a diagram showing a state in which a conditional search is being performed in the first embodiment. [Figure 4] A figure showing the state in which a file related to a patent number list is loaded when performing a number search in the first embodiment. [Diagram 5] FIG. 11 is a diagram showing a search result list in the first embodiment. [Figure 6] FIG. 11 is a diagram showing a state in which information on an individual patent invention is being viewed among the patent inventions displayed in a search result list in the first embodiment. [Figure 7] FIG. 11 is a diagram showing a state in which information on an individual patent invention is being viewed among the patent inventions displayed in a search result list in the first embodiment. [Figure 8] FIG. 2 is a diagram showing a state in which user account information is registered in the first embodiment. [Figure 9] FIG. 13 is a diagram showing a state in which a member menu is selected in the first embodiment. [Figure 10] FIG. 11 is a diagram showing a state in which organization information is set in the first embodiment. [Figure 11] FIG. 11 is a diagram showing a state in which a page for organization settings that has been set is being viewed in the first embodiment. [Figure 12] FIG. 11 is a diagram showing a state in which a page for group settings that has been set is being viewed in the first embodiment. [Figure 13] FIG. 11 is a diagram showing a state in which a new group is being created in the first embodiment. [Figure 14] FIG. 11 is a diagram showing points to note when creating a new group in the first embodiment. [Figure 15] FIG. 13 is a diagram showing a state in which a list of information related to patent inventions placed in a folder in an organizational community is being viewed in the first embodiment. [Figure 16] FIG. 13 is a diagram showing a state in which information on individual patent inventions sorted into folders of an organizational community is being viewed in the first embodiment. [Figure 17] This is a correlation diagram of the patent family related to the patent invention being viewed in the first embodiment. [Figure 18] FIG. 4 is a diagram showing a setting screen for a report mail system in the first embodiment. [Figure 19] FIG. 13 is a diagram showing a report email in the first embodiment for reporting that a patent publication matching a specified condition has been published. [Figure 20] FIG. 11 is a diagram showing a state in which a report email is being viewed in a search system in the first embodiment. [Figure 21] FIG. 13 is a diagram showing a state in which a ranking of notable patents is being viewed in the second embodiment. [Figure 22] FIG. 13 is a diagram showing a state in which a notable poster ranking is being viewed in the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0009] [First embodiment] A search system 1 according to an embodiment of the present invention will be described with reference to Fig. 1 to Fig. 20. Note that the embodiment described below is merely one example of the present invention, and the embodiment of the present invention can be modified as appropriate without departing from the gist of the present invention.
[0010] 1, a search system 1 (including a report mail system) of the present invention includes one or more servers 2, one or more document DBs (Data Bases) 3, and one or more terminal devices 4. The servers 2, document DBs 3, and terminal devices 4 are connected by a network 5.
[0011] Based on user operation, server 2 searches for documents that are the target of the search (in this embodiment, these are the inventions described in the claims and specifications of patent publications and / or published patent publications, as well as the AI abstracts described below; hereinafter referred to as "target documents") and displays the relevant target documents.
[0012] The document DB 3 is a DB that stores target documents. In this embodiment, the document DB 3 is a DB prepared by the business entity that provides the search system 1, but it may be a DB provided by the patent office of each country and / or WIPO (World Intellectual Property Organization), or may be a DB that can be linked to the DB provided by the patent office of each country and / or WIPO via an API (Application Programming Interface).
[0013] The terminal device 4 is a device used by a user who is trying to search for a target document. The terminal device 4 may be a personal computer such as a laptop computer or a desktop computer, a smartphone, or a tablet terminal device. In this embodiment, the terminal device 4 is assumed to be a laptop computer.
[0014] For example, as shown in Figures 2 to 4, the terminal device 4 transmits information capable of identifying the target document (in this embodiment, for example, the application number, patent number, publication number, a conditional expression for searching for patent numbers, a file related to the list of publication numbers, etc.) to the server 2 based on a user operation (input).
[0015] The terminal device 4 receives information related to the target document (for example, the AI abstract, claims, specification, drawings, etc. described below related to the patent number that was the subject of the search) from the document DB 3 via the server 2 and displays it on the monitor 460.
[0016] <Detailed description of Server 2> 1, the server 2 includes a control unit 20, a storage unit 21, and a network interface 22. This server 2 is provided by the business entity that provides the search system 1.
[0017] The storage unit 21 and the network interface 22 included in the server 2 are connected to the control unit 20 of the server 2 via a bus 29 .
[0018] The control unit 20 is composed of a CPU (Central Processing Unit), various input / output interfaces, etc., and controls the operation of the server 2. Details will be described later.
[0019] Furthermore, the CPU of the control unit 20 may be configured by a single integrated circuit, or may be configured by a plurality of circuits (integrated circuits and / or discrete circuits, etc.) communicatively connected.
[0020] The storage unit 21 is mainly composed of a hard disk drive (HDD), a random access memory (RAM), a read only memory (ROM), a solid state drive (SSD), etc.
[0021] The storage unit 21 stores, for example, a computer program for controlling the search system 1 of the present invention, data used to search for target documents, user IDs, and the like.
[0022] The network interface 22 is connected to a communication network 5 for transmitting and receiving data between the document DB 3 and / or the terminal device 4 and the server 2 .
[0023] <Explanation of Functions of the Control Unit 20 of the Server 2> The control unit 20 of the server 2 executes a computer program to function as each of the following means and execute each of the following steps.
[0024] <Summary Step Description> The control unit 20 executes the summarization step by functioning as a summarization unit. The process executed by the control unit 20 as the summarization unit will be described below.
[0025] The control unit 20 controls a summarization program incorporating an API using artificial intelligence (AI). In this embodiment, the summarization program is stored in the server 2, but it may be stored in a device (a device other than the server 2) prepared by the business entity that provides the search system 1.
[0026] As a result, the summarization program generates a summary (hereinafter referred to as an "AI summary") for the invention of claim 1 described in the claims for the specified patent number or publication number, and for an independent claim of the same category as the invention of claim 1 (for example, if claim 1 is an invention of a "method of controlling a vehicle," then the independent claim has the same category, "method of controlling a vehicle." Note that a "method of controlling a sensor" is in a different category from a "method of controlling a vehicle"), as shown in Figures 5, 6, and 20.
[0027] This AI summary is sent from the control unit 20 to the document DB3 and stored in the document DB3 in association with the patent number, etc. (for example, in a CSV file with a file name including the patent number, or in a text file).
[0028] <Registration steps> The control unit 20 executes the registration step by functioning as a registration unit. The process executed by the control unit 20 as the registration unit will be described below.
[0029] As shown in FIG. 2 to FIG. 8, the control unit 20 can log the user into the search system 1 by operating "login" based on the user's operation.
[0030] FIG. 8 shows a state in which a user is registering a user account by clicking "Login" and then "New Registration."
[0031] Specifically, the control unit 20 creates a user ID (user identification information) based on user information (e.g., a user ID, user name, user email address, organization name, and password, which may be set arbitrarily) sent from the terminal device 4, as shown in Fig. 8. In this way, the control unit 20 generates a user account. The user ID is stored in the server 2 and the document DB 3.
[0032] In this embodiment, a user can be uniquely identified by combining a user ID with an email address (and / or a password). In order to set a unique user ID, a warning may be displayed if the user inputs a user ID that has already been registered by another user.
[0033] Furthermore, the control unit 20 creates an organization ID (organization identification information) based on organization information (e.g., organization ID, organization name, master user name, member user name, password) set by the master user for registering the organization ID as shown in Figures 9 to 11 based on the operation of the user whose user account is registered. In this way, the control unit 20 generates an organization account. The organization ID is stored in the server 2 and the document DB 3.
[0034] Figure 9 shows the state in which a user has clicked on "Member Menu" and is about to select one of the items on the displayed "Member Menu" (organization settings, group settings, community, etc.).
[0035] Fig. 10 shows that after "Organization Settings" in the "Member Menu" is clicked, "Create Organization" in Fig. 10 is clicked and each piece of information is entered in the "Create Organization" window. In Fig. 10, "RK" is the master user and is creating the organization.
[0036] 10, an invitation email to join the organizational community of the generated organizational account is sent to the email address of each member user (YN, TS in the example of FIG. 10). When each member user responds, the control unit 20 associates the user ID of the member user with the organization ID.
[0037] If an invitation email is sent to a user who does not have a user account at that time, the user can click the link in the invitation email to go to a login screen where the user can create a new account. The user can then join the organization community by logging in after their account has been authenticated.
[0038] Fig. 11 is a diagram showing a state in which an organizational community of "R Industrial Intellectual Property Department" is set based on the operation of RK, who is a master user. On the screen of Fig. 11, a user can change registration information, change a contract plan, change a payment method, delete an account, and the like. In addition, based on the designation of the master user RK, it may be possible to grant authority to set up an organizational community to any of the member users.
[0039] In the following, the master user and the member users may be collectively referred to as "users".
[0040] Furthermore, the control unit 20 can create a group by registering a "folder" indicating a project or a set of users in the organizational community, as shown in FIGS.
[0041] FIG. 12 is a diagram showing a state in which two groups (group IDs: r-01 and r-02) are set in the organizational community of the Intellectual Property Department of R Corporation.
[0042] A user who is registered as "in-house" in the organization ID can create a folder by specifying a member user, as shown in FIG.
[0043] FIG. 13 is a diagram showing a state in which a new group is about to be set after "Create Group" is clicked based on an operation by a master user or a member user of the organizational community.
[0044] This allows users assigned to the same group to view the same list of patented inventions, as shown in Figures 15 and 16. The examples in Figures 15 and 16 show a state in which user "RK" assigned to group ID (r-01) in Figure 12 is viewing the patented inventions (patent numbers) assigned to "Product A" and the list of them.
[0045] In the example of Figure 15, when user "RK" clicks on the folder for product A and a list of patent inventions (patent numbers) assigned to product A is displayed, user "RK" can right-click on information related to an individual patent invention (patent number) to assign information related to that patent to "Caution" for which a folder has been created within the organizational community.
[0046] In the example of Figure 15, the relevant patent invention (patent number 6718**1) has already been assigned to product A, so "Send to product A" is already checked. On the other hand, product B (group ID: r-02) has not been assigned to user "RK" (see Figure 12), so product B is deactivated so that user "RK" cannot check it.
[0047] Also, Figure 14 shows a warning screen that is displayed when a member of a group includes a user who is registered as "outside the company." This function prevents "shared memos" that should only be shared with users within the company from being shared with users outside the company.
[0048] On the other hand, as shown in Fig. 16, if there is no "external" user among the members of the group, the "shared memo (inside and outside the company)" is deactivated. Alternatively, if there is no "external" user among the members of the group, the "shared memo (inside and outside the company)" item may not be displayed.
[0049] Furthermore, the control unit 20 uses the information related to the summary transmitted from the terminal device 4 to generate a summary created by the user (hereinafter referred to as a "user summary") as shown in Figures 6, 7 and 16. In the examples of Figures 6, 7 and 16, the user summary is entered in the "Everyone's Summary" column.
[0050] This user summary is composed of, for example, a sentence of 300 characters or less in Japanese, paragraph numbers, and drawing numbers. The user summary is also transmitted from the control unit 20 to the document DB 3 and stored in the document DB 3. If the user sets a URL when posting the user summary, a link to that URL is set in the user ID.
[0051] FIG. 7 is a diagram showing the state in which the screen in FIG. 6 has been further shifted (scrolled) downward.
[0052] Also, in the example of Figure 16, the "Everyone's Summary" displayed in the organizational community only displays the user summary that has received the most "likes" (the summary by "XYZ Patent Office" in Figure 6), but other user summaries listed in the "Everyone's Summary" (see Figure 7) may also be displayed.
[0053] Furthermore, the control unit 20 generates tags (in the example of FIG. 6, "vehicle", "carbon dioxide", and "filter") arbitrarily created by the user, as shown in FIG. 6, using the information related to the tags transmitted from the terminal device 4. The tags are transmitted from the control unit 20 to the document DB 3 and stored in the document DB 3. Up to 10 tags can be registered for one patent number. The tags can also be set as conditions for the conditional search in FIG. 3. Furthermore, when the user clicks on a tag (for example, the "vehicle" tag), all patent inventions (patent numbers) to which the "vehicle" tag is registered may be displayed in a list or the like. Furthermore, by checking the text of each tag against a dictionary stored in a predetermined dictionary server, for example, "sensor" and "sensor" may be recognized as the same character, and "AI" and "Ai" may be recognized as the same character.
[0054] In this way, in the search system 1 of this embodiment, the "everyone's summary" and "tag editing" features allow users to communicate with each other within the search system 1.
[0055] Furthermore, the control unit 20 performs settings related to the office advertisement or the recruitment advertisement in Fig. 2 to Fig. 4 using advertiser information (for example, organization name, URL of the organization's homepage, credit card information), recruitment information, etc. transmitted from the terminal device 5. The user can view the recruitment information by, for example, clicking "Recruitment Advertisement" on the menu bar in Fig. 2.
[0056] In addition, the control unit 20 provides a column page using information on a column (for example, a commentary column on patent litigation by a lawyer) transmitted from the terminal device 4. The user can view the column by, for example, clicking "Column" on the menu bar in FIG.
[0057] 9, 18 and 19, the control unit 20 sets conditions for the report mail system using information on a conditional expression (hereinafter referred to as "search conditions") transmitted from the terminal device 4.
[0058] In order for a user to set conditions for the report email system, an organization ID must be created, the user's user ID must be associated with that organization ID, and the destination email address must be registered.
[0059] When these registrations have been made, the user can set the patent owner / applicant (e.g., "XX△ Motors Co., Ltd."), keywords contained in the claims (e.g., "exhaust gas," "filter"), and report frequency (in the example of FIG. 18, "each time," "every three days," "weekly (once a week)," "every other week (once every two weeks)," "monthly (once a month)") that the user would like to know when a new patent publication / publication publication is issued. Note that an arbitrary timing, such as "every Monday," may also be set based on the user's operation.
[0060] 6 and 16, the control unit 20 registers license permission information (e.g., organization name, information indicating that license is permitted) transmitted from the terminal device 5 in association with information on each patented invention (patent number). In this case, for example, a business providing the search system 1 may be able to register each patented invention (patent number) by confirming a certificate transmitted by the user indicating that the user is a legitimate patent holder, or a legitimate patent holder may be able to register each patented invention (patent number) from a setting screen of the organizational community.
[0061] <Authentication step description> The control unit 20 executes the authentication step by functioning as an authentication unit. The process executed by the control unit 20 as the authentication unit will be described below.
[0062] The control unit 20 uses the user ID transmitted from the terminal device 4 to authenticate the user account.
[0063] Specifically, a user can log in with a user name as shown in Figures 8 and 9 by operating "Login" shown in Figures 2 to 8 and entering a user ID and password (an arbitrary password set by the user).
[0064] If each function has been configured, the logged-in user will be able to access the report mail system, access the organization community, and post to the user summary (including "liking" the user summary). Note that if the user only wishes to use the search system 1 (for example, to view the screens in Figures 2 to 7, or to search for patent inventions), the user does not need to go through authentication of the user account.
[0065] In this embodiment, a user with a user ID associated with any of the organization IDs can use the report mail system, but a user with a user ID that is not associated with any of the organization IDs may also be able to use the report mail system by registering to use the system.
[0066] <Reporting step description> The control unit 20 executes the reporting step by functioning as a reporting unit. The process executed by the control unit 20 as a reporting unit will be described below.
[0067] When a patent publication / publication publication that matches the search conditions set in the registration step is issued, the control unit 20 notifies the relevant user of the AI summaries of those inventions (sends them to the email address associated with the organization ID), as shown in Fig. 19. At this time, the control unit 20 provides list data such as "Condition 1.csv" as well as a link for viewing in the report email system in the member menu in the search system 1 (when the user operates the link, the link connects to the report email system).
[0068] In the example of Figure 18, the report email is set to be delivered to the mailing list address (patentgroup@r-kogyo.com) associated with the email addresses of users "RK," "YN," and "TS." However, the report email may be set to be delivered to each user's email address, such as "rk@rk-kogyo.com."
[0069] The control unit 20 obtains an AI summary corresponding to a sentence (e.g., claim 1) related to a patent document extracted by the search conditions at a predetermined timing (e.g., when a new patent publication is published) (causes the summarization program to generate an AI summary).
[0070] Next, the control unit 20 transmits to the terminal device 4 a set of information relating to sentences related to the patent document that match the search criteria (e.g., claim 1 and independent claims in the same category as claim 1) and information relating to the corresponding AI summary, in accordance with the date and time information set based on the user's operation.
[0071] This allows the user to compare the text of the patent document and the corresponding AI summary on the same page by displaying them side by side on a list on the monitor 460 (for example, not limited to one patent invention / patent number, but multiple patent inventions / patent numbers may be displayed vertically in order). Alternatively, the text of the patent document and the corresponding AI summary may be displayed in a separate window.
[0072] <Explanation of measurement steps> The control unit 20 executes the measurement step by functioning as a measurement unit. The process executed by the control unit 20 as the measurement unit will be described below.
[0073] When the control unit 20 detects that the AI summary (including the claims, specification, drawings, etc. relating to the relevant patent number / publication number) has been viewed by a user (or that the patent number has been displayed in a search result list, or that the patent number has been the subject of a report in the report email system), it counts the number of times this has happened. This displays the number of views indicating that the screen relating to that patent number / publication number has been viewed. Note that the examples in Figures 5 and 6 show that the information on Patent No. 6718**1 (the patent number displayed at the top in Figure 5) has been viewed 350 times.
[0074] In addition, the control unit 20 counts the number of times that the AI abstract and user abstract of each patent have been "liked" based on the operation of the logged-in user. The counted number is stored in the document DB 3 in association with each AI abstract and each user abstract. In the example of Fig. 6, the abstract posted by "XYZ Patent Office" has been "liked" 315 times.
[0075] Although not shown, a "like" ranking and / or a "view" ranking may be displayed on the search system 1. In this case, for example, a ranking of users who have created user summaries that have received many "likes" may be displayed, or a list of patent numbers for each patent invention may be displayed in descending order of the number of views. In addition, in the second embodiment described later, a notable patent ranking and notable contributor ranking derived from this will be described.
[0076] Although a user who is not logged in may also be able to "like" a post, it is preferable that only logged-in users be able to "like" a post in order to prevent so-called trolling.
[0077] <Switching step explanation> The control unit 20 executes the switching step by functioning as a switching unit. The process executed by the control unit 20 as the switching unit will be described below.
[0078] The control unit 20 sets whether to display the patent publication or the unexamined publication based on the user's operation.
[0079] Specifically, when both a patent gazette and a public publication have been published for the invention being searched for, the user can switch between displaying the patent gazette or the public publication by operating a button (in Figure 6, "Patent Gazette" is selected from "Patent Gazette / Public Publication").
[0080] Furthermore, when the user operates the "PF" button (see Fig. 6 and Fig. 16), which indicates that the application relating to the patent invention (patent number) has a patent family, the control unit 20 generates information for displaying a correlation diagram of the patent family of the application relating to the relevant patent invention (patent number) on the terminal device 4, as shown in Fig. 17. This allows the user to understand that Japanese divisional applications, U.S. applications, Chinese applications, etc. are associated with Japanese Patent No. 6718**1, and to view summaries in Japanese (AI summary and / or user summary) of each patent invention.
[0081] The control unit 20 can also accept a user operation regarding which of the AI summary and the user summary is to be displayed in the correlation diagram of the patent family.
[0082] Furthermore, the control unit 20 can also display the user summary at the position where the AI summary is displayed in Fig. 6, for example, and display the AI summary at the position where the user summary is displayed in Fig. 6, based on a specified operation by the user (switching the AI summary and the user summary, or erasing the AI summary).This makes it possible to switch the display positions of the AI summary and the user summary (or display only the user summary instead of the AI summary) when the user determines that the user summary is easier to understand than the AI summary.
[0083] In addition to or instead of the above, the control unit 20 may, for example, compare the number of user "likes" (thumbs up) on the AI summary with the number of user "likes" on the user summary, and if the number of "likes" on the user summary is greater, swap the AI summary with the user summary (or display only the user summary instead of the AI summary).
[0084] <Communication step description> The control unit 20 executes communication steps by functioning as a communication unit. The following describes the process executed by the control unit 20 as a communication unit.
[0085] The control unit 20 receives from the terminal device 4 information capable of identifying the target document (e.g., patent number / publication number, patent owner / applicant, application date), user ID, organization ID, email address, password, information related to the user summary, information indicating that a "like" has been given, etc.
[0086] Furthermore, the control unit 20 acquires information related to the relevant patent from the document DB 3, and transmits information capable of identifying the target document (for example, information related to the AI abstract, user abstract, claims, specification, number of views, patent family, etc.) to the terminal device 4. As a result, the screens related to the search system 1 as shown in Figs. 2 to 20 are displayed on the user's terminal device 4.
[0087] <Detailed explanation of document DB3> The document DB 3 is prepared by a business entity that provides the search system 1. In this embodiment, the document DB 3 is prepared separately from the server 2, but may be incorporated in the server 2.
[0088] The document DB3 stores, for example, the following data: ·AI summary Number of "thumbs up" and "thumbs down" responses to the AI summary Number of views of each invention (the total number of times that one or more users of the search system 1 viewed the information on the invention to be viewed) User summary (including paragraph numbers, etc.) - Number of likes for each user summary -Licensing availability Data published in patent publications / publications (bibliographical information such as patent owner and agent information, scope of patent claims, drawings, etc.) User ID User information associated with the user ID (e.g., user name, email address, password, etc.) ·Organization ID Organization information associated with the organization ID (e.g., organization name, user name associated with the organization ID, email address, password, etc.) Group ID Group information associated with the group ID (e.g., folder name, user name associated with the group ID, internal / external information) - User ID assigned to the folder, information related to the invention, etc. - Information on patent families (AI summaries, user summaries, and patent numbers of the patent inventions in the patent family of the patent number being viewed) Advertising information (office information, job information, website link information, etc.)
[0089] The document DB 3 transmits necessary data to the server 2 via the network interface 32 in response to a request from the server 2 (or the terminal device 4 via the server 2).
[0090] <Detailed Description of Terminal Device 4> As shown in FIG. 1, the terminal device 4 includes a control unit 40, a storage unit 41, a network interface 42, a photographing unit 43, a voice input unit 44, an audio processing unit 45, a graphics processing unit 46, and an operation unit 47.
[0091] The memory unit 41 , the network interface 42 , the image capturing unit 43 , the voice input unit 44 , the audio processing unit 45 , the graphic processing unit 46 , and the operation unit 47 are connected to the control unit 40 of the terminal device 4 via a bus 49 .
[0092] The control unit 40 controls the operation of the terminal device 4 based on the user's operation.
[0093] The storage unit 41 is mainly composed of a HDD, a RAM, and a ROM.
[0094] The network interface 42 transmits and receives data between the server 2 and the terminal device 4 .
[0095] The photographing unit 43 is connected to a camera 430. This camera 430 captures an image of the user.
[0096] The voice input unit 44 is connected to a microphone 440. The microphone 440 picks up the user's voice.
[0097] The audio processing unit 45 is connected to a speaker 450. The speaker 450 outputs sound and the like.
[0098] The graphics processor 46 is connected to a monitor 460. The monitor 460 displays a browser screen displayed by the search system 1 and the like.
[0099] The operation unit 47 is connected to a keyboard and a mouse (hereinafter, referred to as "keyboard, etc. 470"). In this embodiment, an operation signal from a user is input to the operation unit 47 via the keyboard, etc. 470, which is an input detection device.
[0100] By operating a keyboard or the like 470, a user can input search criteria such as application number, patent number, publication number, patent classification (IPC, CPC, FI, F-term), keywords, tags (text data set arbitrarily by the user on search system 1), and internal classification (text data set arbitrarily by the user in the organizational community) on screens such as those shown in Figures 2 to 4.
[0101] <Functional configuration of the control unit 40 of the terminal device 4> The control unit 40 of the terminal device 4 executes a predetermined program to function as each of the following means and execute each of the following steps.
[0102] <Input step description> The control unit 40 executes the input step by functioning as an input unit. The process executed by the control unit 40 as an input unit will be described below.
[0103] As shown in Figures 2 to 4, the control unit 40 allows the user to input information about the invention that the user wants to know about (e.g., patent number, keywords) to the search system 1 based on the user's operation, and thereby allows the user to view information about the invention (including a list of inventions) in the search system 1.
[0104] Specifically, the user can input number information such as application number, patent number, publication number, etc., into the window of the search screen for number search shown in Fig. 2. This allows the user to search for the desired target document. Fig. 2 shows the state in which the user has clicked "Number Search", then inputted a patent number (in the example of Fig. 2, "7777777" or "7765363") into the window and is about to click "Search".
[0105] Also, the user can input information such as "CL=gas sensor + vehicle", "TX=exhaust + gas", "FI=B60L*" into the window of the search screen for conditional search shown in Figure 3. This also allows the user to search and view the desired target document. Figure 3 shows the state in which the user has clicked "Conditional Search", then inputted a conditional expression such as a keyword into the window and is about to click "Search".
[0106] Among these, "CL" means the entire scope of claims, "TX" means the entire scope of claims and specification, "FI" means file index, which is a patent classification, "*" means arbitrary ("*" can be any character and any number of characters. Here, it means "forward match"), and "ST" means the life or death information of the application (for example, "0" means extinguished rights, deemed withdrawal due to lack of request for examination, etc., and "1" means existing rights, under examination, etc.). In other words, for example, a user can search for information on patent inventions in which "gasoline" or "vehicle" is mentioned in any of the claims by searching "CL=gasoline+vehicle".
[0107] In addition, the user can input table data that lists patent numbers, etc., into the window of the search screen for number search shown in Figure 4. This also allows the user to search and view the desired target document. Figure 3 shows the state in which the user clicks "Number Search", then drags and drops table data containing multiple patent numbers into the window and is about to click "Search."
[0108] Furthermore, the user can input user information (eg, user name, user address, company, affiliation), user ID, organization ID, and password, and register this information in the server 2 and document DB 3 .
[0109] Furthermore, when the user inputs a correct user ID and password, the server 2 authenticates the user account with that user ID. If the account is authenticated, the user will be able to post user summaries and "like" AI summaries and user summaries, as shown in Figure 7.
[0110] Also, if the user's user ID is associated with one specific organization ID, the server 2 authenticates the organization account using the user ID (the user ID associated with the organization ID). This allows the user to access the organization community screen as shown in Figures 11 to 16.
[0111] Also, the user (master user) can set a user ID to be associated with an organization ID, as shown in FIG. 10. Specifically, the user (master user "RK" in this case) can associate user IDs "RK", "YN", and "TS" with an organization ID of "r-kogyo", for example, as shown in the example of FIG. 10. In the example of FIG. 10, "RK", "TS", and "YN" are set as "in-house" in the organization community of "r-kogyo". In the search system 1 of this embodiment, the example is not limited to that of FIG. 10, and the user can also associate three or more user IDs "in-house" with one organization ID.
[0112] Also, users who can be associated with an organization ID may be limited to "internal" users. In this case, the setting items "internal" and "external" are not displayed. However, even in this case, a user can also associate "external" users with a group ID.
[0113] As shown in Figure 16, users can also enter "shared memos" (text data shared by all users who are set as "internal" in the organizational community), "shared memos (internal and external)" (text data shared across "internal" and "external" in the organizational community), "personal memos" (text data that can only be viewed by users in the organizational community who have registered the "personal memos"), and internal classifications (any text data, such as "filter," "carbon dioxide," "R," "S1," etc.).
[0114] In this way, by using the "shared memo" and "shared memo (inside and outside the company)" of the organizational community provided by the search system 1 of this embodiment, users can communicate with each other. In addition to this, "tags" entered by other users may be displayed on the organizational community screen (e.g., FIG. 16). In this case, the network between users can be further expanded.
[0115] Furthermore, by dragging and dropping data that is associated with a patent number (or application number, etc.) and "Shared Notes" or "Shared Notes (Internal and External)" on the organizational community (including group) screen (for example, by dragging and dropping CSV data with the necessary information entered into the "Product A" folder in Figure 15), the "Shared Notes" and other information entered in the dragged-and-dropped data can be reflected even if "Shared Notes" have not been entered in the organizational community.
[0116] Furthermore, the user may be able to download all or selected information displayed in the organizational community by clicking on “Download” in the organizational community (not shown in FIG. 16).
[0117] Furthermore, by clicking "Create Group" in Fig. 13, the user can create a group within the organizational community as shown in Fig. 13. This allows users assigned to the same group (TS (in-house), RK (in-house), TI (outside the company), and KO (outside the company) in the example of Fig. 13) to view the same folder (Product C in the example of Fig. 13).
[0118] 13, once a user is assigned "in-house" in an organizational community, the user may be automatically assigned "in-house" when creating a group in the organizational community. Furthermore, when a user ID that is not associated with an organizational ID is entered, the user may be automatically assigned "outside the company."
[0119] Furthermore, the control unit 40 sets whether or not to permit licensing of inventions that can be viewed on the search system 1 (for example, "permitted", "not permitted", "- (no input)") based on the user's operation. Specifically, it is possible to set whether or not to permit licensing of the patented invention on the search system 1 based on the operation of a user with legitimate authority (for example, a user who has been authenticated in advance by the search system 1 or the business providing the search system 1 as being an official user belonging to an organization associated with an organization ID). Information regarding whether or not to permit licensing is transmitted to the document DB 3 via the server 2 and is stored in the document DB 3.
[0120] In addition, the user can input, register, and edit search conditions in the report mail system in Figure 18. The user can also set the timing for receiving notifications of new publications. The method for inputting search conditions in this case is the same as the method for inputting conditional search in Figure 3.
[0121] Also, for example, in Fig. 15 and Fig. 20, if the cursor is hovered over a patent number (in this case, the patent number for "6718**1") and a right click is performed, a speech bubble is displayed and the user can assign the corresponding patent to a folder to which the user is assigned. In this case, the user can assign the corresponding patent to the desired folder by checking the box displayed in the speech bubble. Also, if the user associates search conditions with folders (for example, associates condition 1 with product A), the patent related to condition 1 may be automatically assigned to the folder of product A associated with condition 1.
[0122] In the example of Fig. 15, since user "RK" is not assigned to product B (see Fig. 12), commands related to product B are inactive. On the other hand, in the example of Fig. 20, since user "TS" is assigned to both product A and product B (see Fig. 12), both product A and product B are activated.
[0123] <Output step description> The control unit 40 executes the output step by functioning as an output unit. The process executed by the control unit 40 as the output unit will be described below.
[0124] 2 to 20, the control unit 40 generates information for displaying to the user information described in the corresponding patent publication (or unexamined publication) based on information input based on the user's operation. This allows the user to view information on the patent invention (e.g., patent number, patentee, applicant, AI abstract, claims, specification, drawings, user abstract, whether or not a license is granted, and the number of views of the patent invention) on the user's terminal device 4 (monitor 460).
[0125] Furthermore, when a user associated with a specific organization ID logs in, the control unit 40 acquires necessary information from the document DB 3 via the server 2, and displays a screen relating to the organization community on the terminal device 4 (monitor 460) as shown in Fig. 10 to Fig. 17. On this screen relating to the organization community, for example, as shown in Fig. 16, shared memos, shared memos (inside and outside the company), personal memos, in-house classifications, folders created by the user, etc. are displayed.
[0126] Furthermore, if a function for viewing the report mail system is set for the user's organization ID, the control unit 40 displays the AI summaries of the claims of the target document that match the search conditions based on the user's settings on the terminal device 4 (monitor 460) as shown in Figs. 18 to 20. Specifically, the control unit 40 acquires necessary information from the document DB 3 via the server 2, and displays the search conditions set by the user, the number of inventions that match the search conditions, the claims, the description, the AI summaries of the inventions, and the like on the terminal device 4 (monitor 460). Although not shown, a pre-registered advertisement may be displayed on the mail in Fig. 19.
[0127] In addition, based on a user's operation, for example, when a banner associated with a patent firm in Figures 2 to 4 is clicked, the control unit 40 causes the monitor 460 to display an external homepage prepared by the patent firm (for example, an external homepage introducing the firm).
[0128] Furthermore, the user can, for example, check the box to the left of the patent number in Fig. 5 and then operate "Download CSV" to download information related to the checked document (for example, the patent number, applicant, claims, abstract, AI abstract, user abstract, etc., arbitrarily specified by the user, and / or PDF data such as the patent publication). In this case, the terminal device 4 receives information related to the checked document from the document DB 3 via the server 2. Note that "CSV download" may be enabled only when the user is logged in.
[0129] Furthermore, the user can associate any tag with the information related to the patent number, for example by clicking "Edit Tag" in Fig. 6 and then entering any text data (e.g., "Vehicle"). The tag (text data) set by the user can be used as a search item in a conditional search (see Fig. 3). Also, as mentioned above, when the user clicks on a tag, a list of other patent inventions (patent numbers) that have the same tag as the clicked tag registered is displayed.
[0130] Furthermore, the user can transition to information related to a previous or subsequent document by operating the arrows (triangles) shown on both ends of the screen of the search system 1 shown in Fig. 6. Similarly, the user can transition to the previous or subsequent drawing by operating the arrows (triangles) shown on both ends of an item in a drawing.
[0131] In addition, when a paragraph number written in the user summary is specified based on user operation (in the example of Figure 6, paragraph 0100 is specified as shown at the bottom of the page in Figure 6), the control unit 40 displays paragraph 0100 aligned to the top of the frame in the specification item (the area on the right side of the page in Figure 6).
[0132] Also, although not shown in Figure 6, a highlight bar for the keyword entered in the conditional search (e.g., "gasoline", "exhaust") may be displayed next to the specification, and a highlight indicating that the keyword is displayed (e.g., "gasoline" in blue and "exhaust" in red) may be displayed in the corresponding location of the highlight bar.
[0133] Furthermore, the control unit 40 causes the monitor 460 to display an external homepage prepared by XYZ Patent Office, for example, when a URL associated with "XYZ Patent Office" in Fig. 7 is clicked based on a user's operation. In this way, for example, XYZ Patent Office can post a user summary using a user ID of "XYZ Patent Office" and link the URL to its own homepage (external homepage). This allows XYZ Patent Office to guide the user to its own homepage (external homepage).
[0134] <Communication step description> The control unit 40 executes the communication step by functioning as a communication unit. The process executed by the control unit 40 as a communication unit will be described below.
[0135] The control unit 40 transmits to the server information that can identify the target document (e.g., patent number / publication number, patentee / applicant, application date, etc.), user ID, password, user abstract, information indicating that the user "liked" the document, etc. This allows the user to use the search system 1 on the terminal device 4.
[0136] Furthermore, the control unit 40 acquires information related to the relevant patent from the document DB 3, and receives information capable of identifying the target document (e.g., the scope of claims, the specification, shared memos, personal memos, in-house classification, information indicating that the relevant invention has been "liked") from the server 2 via the document DB 3. This allows the user to view a screen related to the search system 1 on the terminal device 4.
[0137] In summary, the report mail system of the present invention comprises the following steps: A computer system that causes a computer (a server 2, a document DB 3, a terminal device 4) to execute the following steps: In the input step, based on a user's operation, conditions (e.g., keywords, patent classification) for searching document information (e.g., patent number) of a target patent document, and date and time information (e.g., every time, every three days, once a week, etc.) for specifying the frequency of obtaining summary information (AI summary) corresponding to sentences (e.g., claim 1 and independent claims in the same category as claim 1) related to the document information (e.g., patent number) searched for by the input conditions are input. An AI summary is information that summarizes sentences (e.g., claim 1 and an independent claim in the same category as claim 1) related to document information (e.g., a patent number) using a predetermined learning model, In the acquisition step, an AI summary is acquired separately from the text related to the document information. In the transmission step, information indicating that the AI summary has been acquired is transmitted to an operation terminal 4 operated by a user according to the date and time information; In the output step, the document information is output to the monitor 460 of the operation terminal 4 so that the user can compare the document related to the document information (e.g., claim 1 and independent claims in the same category as claim 1) and the corresponding AI summary on the same page.
[0138] In addition, the search system 1 of the present invention includes: A computer system that causes a computer (a server 2, a document DB 3, a terminal device 4) to execute the following steps: In the input step, the document information (e.g., patent number) of the target patent document or the conditions for searching the document information (e.g., keywords, patent classification) are input based on the user's operation, In the acquisition step, summary information (AI summary) corresponding to the sentences related to the input document information or the document information searched by the conditions (e.g., claim 1 and an independent claim in the same category as claim 1) is acquired, AI summaries are information that summarizes sentences related to document information using a specified learning model. In the output step, summary information that matches the input document information (eg, patent number) or conditions (eg, keywords, patent classification) is output to the monitor 460.
[0139] <Effects of the Invention> As described above, according to the present invention, it is possible to provide functions that have not been provided in conventional patent search systems.
[0140] This makes it possible to provide a computer system that can provide a novel patent search system.
[0141] [Second embodiment] A ranking system (notable patent ranking system, notable contributor ranking system) according to an embodiment of the present invention will be described with reference to Fig. 21 and Fig. 22. Note that the embodiment described below is merely one example of the present invention, and the embodiment of the present invention can be modified as appropriate without changing the gist of the present invention. Also, the symbols used in the second embodiment are the same as those in the first embodiment.
[0142] In this embodiment, a system for creating a notable patent ranking and a notable contributor ranking using the indexes (variables) used in the first embodiment or the indexes (variables) created by the search system 1 will be described.
[0143] First, the notable patent ranking system will be described. When creating the notable patent ranking, the server 2 creates a daily, weekly, monthly, and award (yearly) ranking of notable patents.
[0144] Figure 21 shows an example of a ranking compiled for a specified period by the notable patent ranking system. The following shows an example of a ranking compiled monthly, but similar rankings can also be created daily, weekly, etc.
[0145] In Fig. 21, the rankings of notable patents are tallied for each of sections A to H of the patent classification FI, and the total value of these rankings is displayed as an overall ranking. These "sections" are predefined based on the IPC (International Patent Classification), such as "A" for "Everyday necessities," "B" for "Processing operations; transportation," and "C" for "Chemistry; metallurgy." Note that, although the FI is used as the patent classification in this embodiment, the IPC may also be used.
[0146] In the section-by-section aggregation, variables used are, for example, the number of views, the number of user abstracts posted, the number of "likes" for the user abstracts, the number of allocations to folders, etc. for each patent invention (patent number). The aggregation is for a certain period, for example, from October 1, 2023 to October 31, 2023.
[0147] In this embodiment, for example, the server 2 assigns points to the patented invention (patent number), such as 100 points if the number of views of the target patented invention (patent number) is 100, 10 points if the number of posts is 10, and so on, and accumulates the points assigned to each patented invention.
[0148] Alternatively, smoothed points may be awarded by dividing the number of views of each patent invention (patent number) by the number of views of the patent invention (patent number) with the most views in the same section. Specifically, if the most viewed patent in section "A" is, for example, patent No. 9999999, and the number of views is 10000, server 2 will award 1 (=10000 / 10000) point to patent No. 9999999, and if the number of views of patent No. 9999998 in section "A" is 100, server 2 will award 0.01 (=100 / 10000) point.
[0149] By doing this, even if there is a bias in the number of views between sections (for example, if the number of views varies greatly between sections, such as section "A" having a total of 20,000 views and section "B" having a total of 300 views), the ranking can be determined based on the smoothed points across the sections.
[0150] The server 2 also performs similar tallying for the number of user abstracts posted, the number of "likes" given to the user abstracts, the number of allocations to folders, etc. The number of allocations to folders is the cumulative value when the patent invention is allocated to a "needs attention" folder, etc., for all users.
[0151] The server 2 can also calculate the growth rate of the number of views, user summaries, etc. from the previous tally (for example, if the current tally is between October 1, 2023 and October 31, 2023, the previous tally will be between September 1, 2023 and September 30, 2023). This allows points to be awarded for each patent invention (patent number) according to the growth rate of the number of views, etc.
[0152] The server 2 performs tallying for each section and creates a ranking of the patent inventions (patent numbers) that have received the most attention (awarded points) in each section.
[0153] Then, an overall ranking is created for all patent inventions (patent numbers) across all sections from A to H, in order of the amount of attention they receive.
[0154] In addition, when the "+patent" column in Fig. 21 has a "○", it means that the relevant patent holder has created a patent PR page to promote the patent in the search system 1. This "+patent" contains information such as an outline of the invention described by the patent holder, the product in which the patent invention is implemented, whether or not a license is available, whether or not the patent right can be transferred, and a link to move to the detailed page of each patent invention (for example, the screen in Fig. 6).
[0155] In addition, if the details page of each patent invention, such as that shown in Figure 6, also has a link to "+patent," the patent holder can increase the number of views of the patent invention by linking the details page of the patent invention with "+patent."
[0156] 21, if the text data (patent holder name) in the patent holder column is underlined, it means that each patent holder has created an organization page to promote itself in the search system 1. In this organization page, the patent holder can register "+patent" and announcements (for example, advertisements for new products), etc.
[0157] In this way, patent holders can attract more attention to their patent inventions by creating a "+patent" organizational page.
[0158] Next, the notable contributor ranking system will be described. In creating the notable contributor ranking, the server 2 creates a daily, weekly, monthly, and award (one year) ranking of users (hereinafter referred to as "contributing users") who have posted notable posts (posting user summaries) in the same way as the notable patent ranking.
[0159] FIG. 22 is an example of a ranking compiled on a "weekly" basis by the notable contributor ranking system.
[0160] In the notable contributor ranking, the server 2 assigns points to each posting user in a manner similar to that of the notable patent ranking, based on at least the number of posts by each posting user and the number of "likes" for the user abstract. The server 2 then accumulates the points assigned to each posting user, and arranges the posting users in order of the points assigned. Note that, like the notable patent ranking, points may be assigned based on the growth rate of the number of posts and / or the number of "likes" from the previous count.
[0161] Furthermore, if the "HP" column in Figure 22 contains a "○", this means that the search system 1 contains a link to the external homepage of the corresponding posting user (e.g., a patent attorney, a patent attorney corporation, a patent holder, or an employee of a corporate intellectual property department).
[0162] Also, in Figure 22, if the organization name in the "Organization" column is underlined, it means that each posting user (e.g., patent attorney, patent attorney corporation, patent owner, employee of a corporate intellectual property department) has created an organization page in the search system 1 to promote themselves (an internal page that differs from the user's external homepage and includes, for example, a list of specifications created by the posting user, the posting user's own patent allowance rate, rejection rate under Article 36 of the Patent Act, etc.).
[0163] In this way, by posting a link to their own external homepage or by creating an organization page within the search system 1, the posting user can attract even more attention to their own existence.
[0164] The ranking information regarding the organization page, "+patent", notable patent rankings, and notable contributor rankings is transmitted from the server 2 to the document DB 3 and stored in the document DB 3. The server 2 transmits this information to the terminal device 4 in response to a user's operation. As a result, the notable patent rankings and notable contributor rankings are displayed on the terminal device 4.
[0165] In the notable patent ranking, variables such as the number of views of each patent invention (patent number), the number of user abstracts posted, the number of "likes" for the user abstracts, and the number of allocations to folders are used, but these variables may be weighted in a predetermined manner. For example, if the number of allocations to folders is weighted more than the number of views, even if the number of views is 1 and the number of allocations to folders is 1, 1 point is given to the number of views and 2 points are given to the number of allocations to folders, and different points are given for each variable.
[0166] Also, in the notable contributor ranking, unlike the notable patent ranking, the server 2 does not create a ranking of posting users for each FI, but can evaluate posting users who actively post user summaries to patent inventions (patent numbers) with FIs with a small number of posts as a variable for awarding points. In this case, for example, if the number of user abstracts posted to section "A" is smaller than the other sections, the server 2 can award three points, rather than one point, to posting users who posted user summaries to section "A".
[0167] Also, while Figures 21 and 22 show examples in which the user viewing the rankings is logged in, users may be able to view the notable patent rankings and notable contributor rankings regardless of whether they are logged in or not.
[0168] Furthermore, the variables and calculation methods used to calculate the points for the notable patent rankings and notable contributor rankings are not limited to the above examples, and various variables and calculation methods can be used.
[0169] In summary, the ranking system of the present invention is as follows: A computer system that causes a computer (a server 2, a document DB 3, a terminal device 4) to execute the following steps: In the input step, document information (e.g., patent number) of a target patent document and / or conditions for searching document information (e.g., keywords, patent classification) are input based on a user's operation; In the generating step, viewing information (e.g., the number of views, the user summary, the number of posts of the user summary, the "likes" for the user summary, and the number of "likes") is generated regarding the viewing of the text related to the document information by the user; In the output step, at least the document information and the viewing information are output to an operation terminal 4 operated by a user. In the calculation step, an index relating to the attention of the patent document (a notable patent ranking, a notable contributor ranking) is calculated at a predetermined timing (e.g., daily or monthly), The index relating to the attention level of a patent document is an index calculated based on the browsing information for each patent document or each user.
[0170] <Effects of the Invention> Previously, there was no system that created rankings based on the number of views of individual patent inventions (patent numbers), etc.
[0171] In contrast, according to the present invention, highly popular patents and highly popular posting users are displayed by ranking, which allows users to view patent inventions not only from a technical perspective, but also from the perspective of which patent inventions are frequently viewed by users.
[0172] [Other embodiments] The various control means and processing procedures described in the above embodiments are merely examples, and are not intended to limit the scope of the present invention, its applications, or its uses. The various control means and processing procedures can be appropriately modified in design without changing the gist of the present invention. In addition, the various control means, processing procedures, and processing contents described in each embodiment can be configured individually or in appropriate combinations.
[0173] In addition, the various definitions used in this specification are commonly used throughout the specification unless otherwise specified in each example.
[0174] In the above embodiment, the search system is configured with one server and one document DB, but the present invention is not limited to this. The search system may include two or more servers and two or more document DBs, and the document DB may be stored in the server.
[0175] In the above embodiment, an example is described in which the claims are read in advance into the summarization program stored in the server to create an AI summary, but the present invention is not limited to this. For example, when a user inputs a patent number related to a target document, an AI summary of the invention related to the claims of the corresponding patent number may be created in real time by a summarization program stored in the server (or stored in an external device that receives instruction information from the server).
[0176] In the above embodiment, an example is described in which three user IDs are associated with one organization ID, but the present invention is not limited to this. For example, four or more user IDs may be associated with one organization ID. Also, for example, a user may be able to associate five or more user IDs with one organization ID by paying a fee (for example, 1,000 yen / ID) to a business providing a search system.
[0177] In addition, by paying a fee, a user may be able to use a report mail system, add folders within a group, view a correlation diagram of patent families, set "outside the company," and so on.
[0178] Furthermore, without being limited to the above embodiment, among the information stored in the document DB, for example, shared memos, shared memos (inside and outside the company), personal memos, company classifications, etc. may be stored only in the terminal device of each user. In this case, the security of the search system is further improved.
[0179] In addition, in the above embodiment, an example is described in which a Japanese patent number / publication number is input, but the present invention is not limited to this. When a patent number in each country such as Japan, the United States, or China (for example, 6718**1jp (or jp6718**1, the same applies below), 7******us, or 5******cn) is input, the search results may be displayed in the same format as in the above embodiment.
[0180] In addition, information on a patent family can be specified by "application number" and / or "patent number," as well as "country type," etc. For example, if "PCT / US application B" is listed as the priority claim number in the bibliographic information of "Japanese application A," it is uniquely determined that application A and "U.S. application B (or U.S. application C transferred to the U.S. based on a PCT application)" related to "PCT / US application B" are a patent family.
[0181] Also, when the base application is "application a" and its divisional application (child application) is "application b", it may be possible to identify that "application a" and "application b" are a patent family based on the original application "application a" listed in the bibliographic information of the child application "application b". Also, when the server searches the document DB for an application based on "application a", and "application a" is listed as the base application in the bibliographic information of the detected "application b", it is identified that "application a" and "application b" are a patent family. Also, the same patent family number may be assigned to the same patent family.
[0182] Furthermore, when a list of search results is displayed on a terminal device, the patented inventions (patent numbers) may be displayed in order of the number of views (or in order of application date or patent number if the number of views is the same). Furthermore, for each patented invention (patent number), the user summaries may be displayed in order of the number of "likes" given to the user summaries. Furthermore, the user summaries may be displayed grouped by claim, such as an area for displaying claim 1, an area for displaying claim 5, and so on.
[0183] In addition, in the document DB, shared memos, shared memos (inside and outside the company), personal memos, etc. may be recorded in an encrypted state. When a user views a shared memo, the server acquires information such as the encrypted shared memo, and transmits it to the terminal device either in the encrypted state (in this case, the information such as the shared memo is decrypted by the terminal device) or in the decrypted state.
[0184] Furthermore, for user IDs associated with an organization ID, "inside the company" may be automatically assigned when a group is created. On the other hand, for user IDs not associated with an organization ID, "outside the company" may be automatically assigned when a group is created. This makes it possible to prevent an "outside the company" user from being erroneously assigned to "inside the company".
[0185] Additionally, the AI summary and / or the user summary may be configured to not allow copying and pasting of text data, and this feature may be applied only to users who are not logged in.
[0186] In addition, in the search system, display of search results and / or downloading of CSV files may be permitted as long as it does not exceed a predetermined upper limit.
[0187] In addition, by having the AI learn based on the user's specifications or the user's settings, when a publication summary that matches the conditions is reported in the reporting email system, information related to each invention may be automatically assigned to a folder within the organizational community (e.g., Patent No. 6717*** is Product A, Patent No. 6718**1 is Product B), or an internal classification (e.g., Patent No. 6717*** is "Autonomous Driving", Patent No. 6718**1 is "Purification Filter").
[0188] In addition, when an advertiser sets a patent classification (e.g., FI, IPC) of his / her choice, if the patent classification associated with the patent invention searched by the user is the same as the patent classification set by the advertiser, the advertisement information of the advertiser may be displayed in the search results. In this case, if the person in charge of patent office A specifies, for example, "FI=A63F / *" as the technical field in which patent office A excels, when "FI=A63F / *" is associated as the top FI with the patent invention searched by the user, the advertisement of patent office A will be displayed on the details page of the extracted patent invention.
[0189] In the above embodiment, an example is described in which the report mail system reports information on patent publications that match the conditional formula to the user, but the present invention is not limited to this. That is, the report mail system may also be configured to notify the user of information on publications that match the conditional formula (and their AI summaries). In addition, in the report mail system, the CSV file attached to the mail in FIG. 19 may include at least one of the patent owner, claims, detailed description of the invention, and representative diagram, in addition to the patent number and AI summary. That is, for example, the file may display at least the patent number and the AI summary of all independent claims (otherwise, it may be only claim 1 or all claims).
[0190] Also, although not described in the above embodiment, multiple users (computers) may be able to log in simultaneously to one user ID, but it is preferable that only one user (one computer) be able to log in to one user ID.
[0191] In addition, if a user has linked their user ID to a social networking service (SNS), when that user posts to everyone's summary, that summary may be posted on the SNS.
[0192] The search system may also provide a function that allows message chat within the organization on the organizational community screen, and may provide a function that notifies other organization members when any user member updates a "shared memo" or checks information about a specific patent.
[0193] In addition, in the above embodiment, an example is described in which one AI summary is generated (and / or displayed) corresponding to one claim (e.g., claim 1) of each patent number, but the present invention is not limited to this. For example, for claim 1, multiple types of AI summaries may be generated (and / or displayed), such as a one-word AI summary of 20 characters or less (e.g., for quick checking), an AI summary of 100 characters or less (e.g., for development department personnel), and a detailed AI summary of 200 characters or less (e.g., for patent department personnel). In this case, for example, the user may be able to select at least one of the "one-word AI summary of 20 characters or less", "AI summary of 100 characters or less", and "detailed AI summary of 200 characters or less".
[0194] As described above, the specification that allows the user to select at least one of "one-word AI summary of 20 characters or less", "AI summary of 100 characters or less", and "detailed AI summary of 200 characters or less" can also be adopted in the report mail system. In this case, it is sufficient that the screen in Fig. 18 is provided with a button for selecting at least one of "one-word AI summary of 20 characters or less", "AI summary of 100 characters or less", and "detailed AI summary of 200 characters or less".
[0195] For example, when "AI Summary of 20 characters or less" and "AI Summary of 100 characters or less" are selected, the user can compare the "AI Summary of 20 characters or less" and "AI Summary of 100 characters or less" of claim 1 (and other independent claims) with claim 1 (and other independent claims). This allows the user to more easily understand the inventions related to the claims.
[0196] Furthermore, the "multiple types of AI overviews" may be generated using the same learning model, or each may be generated using a different learning model.
[0197] The AI summary displayed on the system may be claim 1 and the independent claims of the same category as claim 1, or may be all the independent claims, or may be all the claims (including dependent claims). When extracting only the independent claims, for example, the following procedure may be executed.
[0198] For example, a summarizing program is made to read the scope of the patent claims to be searched, and claims that contain the word "claim" in the claim text (for example, claims that contain "the program described in claim 1" in the claim text) are excluded. The remaining claims are then treated as independent claims.
[0199] Furthermore, when extracting only independent claims in the same category as claim 1, the abstract program identifies the category of claim 1 (for example, "program"). An example of the identification method is shown below.
[0200] For example, the summarizing program first extracts the last character (e.g., "mu") of claim 1 and judges whether the character (mu) is included in the title of the invention in the bibliographic information (e.g., "control method, program"). If the title of the invention contains only the character (mu) (in the above example, only "program" contains "mu" and "control method" does not), the independent claims whose last character is the character (mu) are judged to be independent claims in the same category as claim 1.
[0201] On the other hand, if the title of the invention contains the character (mu) multiple times (for example, if the title of the invention is "Control method, program, system"), the summarization program extracts the last two characters of claim 1 ("ram"), and if the title of the invention contains the character (ram) only once, it determines that the independent claims whose last two characters are the character (ram) are in the same category as claim 1. If no independent claims in the same category are identified even after this, the scope of confirmation can be expanded by, for example, extracting the last three characters of claim 1.
[0202] In addition, if the last character of claim 1 is not included in the title of the invention, the same procedure as above (e.g., comparison with the title of the invention) may be carried out using the first character before "wherein" or "wherein". If the first character does not identify an independent claim of the same category, the same procedure is carried out using the two characters before "wherein" or "wherein".
[0203] The summarization program reads text such as claim 1 and has the AI API create a summary of it. Furthermore, the summarization program removes unnecessary phrases from the summary such as "This text is," "This patent is," and "Details omitted," and can also perform morphological and syntactic analysis of the sentences in the summary, correcting sentences written in the "desumasu" style to the "dearu" style. The sentences that have undergone these processes are finally generated as the AI summary.
[0204] In addition, in the above embodiment, an example is described in which the control unit controls the summarization program, but the present invention is not limited to this. The summarization program may be managed by an external device of the server, and may be capable of summarizing the wording of the claims related to the patent number to be searched based on an instruction to the external device. The generated AI summary may then be stored in the document DB from the external device.
[0205] In addition to "Job Advertisement" and "Column", the menu bar may also include a "Bulletin Board". In this bulletin board, users may be able to exchange opinions with other users using their account IDs or anonymous IDs.
[0206] In addition, although not shown in Fig. 6, a column for an agent may be provided in addition to the patent number, patent owner, etc. In this case, a link to the agent's external website may be attached to the agent's name or identification number written in the agent's column. This allows a user who has viewed the patented invention (patent number) to easily understand information about the agent by clicking the link and referring to, for example, an external website opened in a separate tab.
[0207] Furthermore, without being limited to the above embodiment, when an organization ID has been created, a patented invention (patent number) may be shared with other users belonging to the organization community in the following format. For example, a user can specify a patented invention (patent number) in the scene of Fig. 15 and right-click to "bookmark" it. With this "bookmark", for example, up to three folders can be created for one user, and a user who has "bookmarked" a patented invention (patent number) can assign that patented invention (patent number) to at least one of the folders.
[0208] Then, by designating one of the user's folders as "shared with organization community," the user can share the folder with other users who belong to the same organization community.
[0209] In addition, in the above embodiment, an example is described in which the target document is a patent document, but the present invention is not limited to this. The target document may be literature such as a novel, a paper (e.g., a technical paper), a design gazette, or a trademark registration gazette. This allows the registered designs and registered trademarks that are attracting attention to be displayed in a ranking, so that the user can easily understand the trademarks of interest.
[0210] In addition, an application that works in conjunction with the search system of the present invention may be provided, and the user may configure the application so that a push notification is sent to the application when a new AI summary is available for notification via the report email system.
[0211] In addition, AI can be trained to do the following: For example, AI can be made to read the specifications (detailed descriptions of the invention) and drawings of all (not only Japanese but also foreign) patent documents (not only patent documents for which an abstract is to be generated, but all other published patent documents) and learn the meanings of terms in all patent documents, thesaurus, etc. Then, AI can be made to read the claims of each individual patent publication and summarize the inventions related to the claims, while explaining the wording of the claims using descriptions in the specifications or drawings, annotating the wording of the claims based on descriptions in the specifications or drawings, or replacing the wording of the claims with descriptions in the specifications or drawings.
[0212] For example, if the claim: “[Claim 1] A computer system that causes a computer to execute the following steps: In the input step, based on a user's operation, a condition for searching document information of a target patent document and date and time information for specifying a frequency of obtaining summary information corresponding to a sentence related to the document information searched for by the input condition are input; The summary information is information obtained by summarizing a sentence related to the document information using a predetermined learning model, In the acquisition step, the summary information is acquired separately from the text related to the document information; In the transmission step, information indicating that the summary information has been acquired is transmitted to an operation terminal operated by the user according to the date and time information; In the output step, the summary information is output to the operation terminal so that the user can compare the sentence related to the document information with the summary information. Computer systems." and the specification states that "document information of a patent document is, for example, a claim," "summary information (AI summary)," and "the frequency can be set, for example, to once a day or once a week," then when a new patent gazette is issued, the summarization program can read the claims, specification, and drawings to generate an AI summary such as "a system that, when a new patent gazette is issued, obtains summary information (AI summary) of the invention related to claim 1 in the patent gazette, and displays claim 1 and the AI summary on a terminal device in accordance with the specified frequency (once a week) so that they can be compared," and this AI summary can be displayed on a computer system.
[0213] This makes it possible to summarize (explain) the claimed invention using wording in the specification that is easier to understand (or that applies or replaces wording in the specification). Also, if the drawings say "output unit 10" and claim 1 says "a system that displays the previous day's data on a display unit," and the summarization program recognizes that "output unit 10" and "display unit" are the same concept, the AI summary may be able to fit the wording on the drawings to the description of the claim, such as "something that displays the previous day's data on the output unit 10" or "something that displays the previous day's data on the display unit (output unit 10)." [Explanation of symbols]
[0214] 1. Search System 2 Server 3 Document DB 4 Terminal Equipment
Claims
1. A computer system that causes a computer to execute the following steps: In the input step, a search query for searching document information of a target patent document can be input based on a user's input operation, and a receiving operation can be received by the user; the receiving operation is an operation for receiving summary information corresponding to a sentence regarding the document information of the newly issued patent document when a condition that the newly issued patent document is satisfied; the summary information is information obtained by summarizing a sentence related to the document information using a predetermined learning model; In the acquisition step, the summary information is acquired separately from the text related to the document information; In the transmission step, when the receiving operation is accepted, information indicating that the summary information has been received is transmitted to an operation terminal operated by the user; In the output step, when the receiving operation is accepted, the summary information is output to the operation terminal so that the user can compare the sentence related to the document information with the summary information. Computer system.
2. the summary information includes a plurality of types of summary information generated according to a plurality of criteria; In the input step, information for acquiring the summary information generated in accordance with at least one of the plurality of criteria is input based on an operation of the user.
2. The computer system of claim 1.
3. The summary information is Of the claims, specification and drawings contained in said patent document, The claim language is explained using the description of the specification and / or drawings, not just the claim language; the description of the specification and / or drawings is used to annotate the claim language; and The language of the claims is replaced with the description in the specification and / or drawings. At least one of the following:
2. The computer system of claim 1.
Citation Information
Patent Citations
Consumer trouble resolution system and method
JP2003114927A
Infringement information extraction system, method, and program
JP2022073872A
Extraction program, extraction method and information processor
JP2023039832A
