Method for managing patents by means of artificial intelligence

Abstract

Method for managing patents by means of artificial intelligence algorithms implemented on a computer system, the method comprising the following steps: - importing patent data from a database in an automatic or manual way, or if the patent does not exist by activating a wizard to draft a patent application, - processing the imported patent data, using artificial intelligence algorithms to generate images and videos of the same and a list of possible fields of application of the technology inherent to the patent data, - searching for a patent through a text / image / field of application search and / or self-generated "use-cases" and / or through the correlation of multiple patents to extract the characteristics of a specific product from these patents, - composing a new patent text to be subsequently submitted to a patent office, in the event of a negative outcome of the search, using a wizard, - transacting a commercial operation, for example, purchase / sale of a patent, acquisition / granting of a license, - identifying groups of collaborators to enhance their IP portfolio and / or patents based on market trends and / or scientific interests.

Description

[0001] METHOD FOR MANAGING PATENTS BY MEANS OF ARTIFICIAL

[0002] INTELLIGENCE

[0003] Technical field of the invention

[0004] The present invention relates to a method for managing patents by means of artificial intelligence. In particular, the system focuses on the dissemination and sale of patents, through an associated platform.

[0005] Background art

[0006] Numerous methods are known for searching and consulting published patent documents. Many of the state-of-the-art databases, some of which are free, allow you to carry out free searches on patent documents. The known methods of archiving and consulting published patent documents use platforms in which it is possible to carry out searches, even combined, by means of keywords contained in the title and / or in the text, the indication of the filing date or a time range, of the names of the inventors and / or applicants, etc. What is provided as a search result is the patent or a list of patent documents corresponding to the search parameters entered. It is also possible to further filter this list with additional parameters.

[0007] These systems and search methods, although efficient, are dependent on the user entering data. There is therefore often the risk of not bringing out potential patents of interest just because the keywords and / or data entered do not correspond to the parameters you want to search for.

[0008] In addition to this, the patent documents that are extracted are generally very numerous and it is therefore further difficult and demanding to be able to identify the most relevant and / or of interest. This therefore requires further and in-depth analyzes of individual patent titles.

[0009] Furthermore, known platforms do not respond to commercial needs which, to date, are managed directly by the patent holders, through lengthy negotiations. The exclusive rights conferred by law on the owner are obtained with the granting of the patent. The patent title can therefore be, depending on the agreements, assigned under an exclusive or nonexclusive license with consequent payment of a royalty, or sold to third parties. The current systems and methods of consulting patent documents do not allow for the creation of a favorable connection for the development of technologies and scientific discoveries, as they do not encourage their application on the market and the exchange of patented technologies. There is therefore a need to define an innovative method and an associated service to facilitate search and transactions related to patent management, which are free from the drawbacks mentioned above.

[0010] Summary of the invention

[0011] An aim of the present invention is to define a method for managing patent documents by means of artificial intelligence (Al) algorithms, in order to also facilitate commercial transactions between patent holders.

[0012] The method is characterized by:

[0013] - insertion of patent data into a database; - acquisition of patent data from the database;

[0014] - processing of said data with the generation of images and videos of the same and a list of possible fields of application of the technology;

[0015] - search for a patent by means of text / image / field of application and / or by correlating multiple patents for the creation of a specific product.

[0016] Advantageously, the method according to the present invention is able to enhance and share the teachings indicated in the patent text by analyzing the patent data by means of artificial intelligence algorithms.

[0017] Advantageously, by means of artificial intelligence, the method is able to generate images and videos starting from the patent description and also allows you to independently calculate and propose all the possible applications of a given technology. The method allows to recover valuable information through the analysis of "big data" and to summarize key information by representing it in a visual way and creating possible applications for use in the real world.

[0018] Advantageously therefore, the method according to the present invention introduces an innovative approach in the purchase, sale and use of patents also applying the concept of subscription and credits to access the platform and use the rights according to your subscription plan.

[0019] A further purpose of the present invention is to provide a secure and confidential environment, for example a computer system or one or more servers, in which buyers and sellers can evaluate patents and successfully conclude transactions with certainty. Advantageously, the secure and confidential environment includes an e-commerce module that exploits artificial intelligence techniques to improve search and analysis processes, through, for example, a graphic representation of the findings described in the patent text.

[0020] Therefore, according to the present invention, a method for managing patent documents is provided, based on artificial intelligence algorithms and having the features set out in the method claim, attached to the present description.

[0021] According to another aspect, a computer system is provided for the implementation of the method for managing patent documents having the features set out in the product claim, attached to this description.

[0022] Further embodiments of the invention, preferred and / or particularly advantageous, are described according to the features set out in the attached dependent claims.

[0023] Brief description of the drawings

[0024] The invention will now be described with reference to the attached drawings, which illustrate some non-limiting implementation examples, in which:

[0025] - figure 1 is a synthetic logical diagram of the patent management method, according to a preferred embodiment of the present invention,

[0026] - figure 2 is a detailed logical diagram of the method in Figure 1,

[0027] - figure 3 is a block diagram of the flow of activities for importing a patent, according to an automatic procedure,

[0028] - figure 4 is a block diagram of the flow of activities for importing a patent, according to a manual procedure - figure 5 is a block diagram of the activity flow of creating a patent application, according to the present invention,

[0029] - figure 6 is a block diagram of the patent search activity flow, according to the present invention,

[0030] - figure 7 is a block diagram of the phases for completing an online transaction, according to the present invention,

[0031] - figure 8 is a block diagram of the activity flow of a user of the method according to the present invention.

[0032] Detailed description

[0033] By way of purely illustrative and non-limiting example, the method 100 for managing patent documents by means of artificial intelligence algorithms, according to the present invention, will now be described with reference to the aforementioned figures 1-8.

[0034] With reference to Figure 1, the method 100 is implemented by means of a software platform resident on a computer system. The software platform includes four main modules:

[0035] - a first module 10 for importing and "storing" data;

[0036] - a second search module 20;

[0037] - a third module 30 of "e-commerce";

[0038] - a fourth "social" module 40;

[0039] - a fifth data processing module 50, based on artificial intelligence algorithms.

[0040] Figure 2 illustrates more specifically the different modules that will be described below also with reference to specific detailed figures.

[0041] According to the first import module 10, the data within the software platform can be inserted via two modes: an automatic insertion mode illustrated in Figure 3 and a manual insertion mode illustrated in Figure 4 and Figure 5.

[0042] The automatic mode involves massive import of data from third- party databases, for example public domain patent databases. With this method it is possible to import all the patents belonging to a specific technological sector, for example electric vehicles, etc. The automatic mode involves the acquisition of patents through search indices which will extract the patents from existing databases and register them on the software platform, classifying them on the basis of the structural data present. After import, the non-structural data will be transformed by the fifth data processing module 50 into useful information for a more complete classification and presentation of the patent compared to the prior art. In particular, for each patent a visual output will be created (images and / or videos) which will represent its field of application, the context of possible use, possible alternative self-generated application use-cases and market trends and sales options.

[0043] The manual mode involves the insertion of a specific patent (Figure 4) or the insertion of data for the creation of a new patent (Figure 5). The manual mode envisages, again after data entry, the creation through the software platform of a series of possible "use-cases" i.e. methods of alternative use of the technology / solution proposed in the patent also creating correlations with other patents present in the database, based on the field of application, the context of possible use and market trends and sales options. An inventor or patent holder can manually insert a patent into the software platform via a specific request form in which the patent file can be uploaded in .pdf format and enriched with all the metadata and tags deemed of interest.

[0044] The identification of alternative "use-cases" includes the following steps:

[0045] -analyze similarity: Al-based algorithms compare the entered patent with other similar inventions to identify uses that are not immediately obvious, -scenario simulation: uses simulation models to predict how the invention could be adapted to different application contexts, -data-based suggestions: provide suggestions based on historical data of use of similar technologies.

[0046] With reference to Figure 5, in the event that a patent does not yet exist, through specific registration within the software platform, the user can use a special "wizard" (guided procedure) that accompanies him step- by-step to draw up a draft pre-compiled patent application, and obtain a complete and structured document.

[0047] Specifically, the wizard for drafting a patent text includes specific fields to be filled in by the user, accompanied by suggestions and detailed explanations. The fields to fill in are as follows: a. Title of the Invention: the user enters a descriptive and concise title of the invention; b. Description of the Invention: this field requires a detailed description of the invention, including the technical problem solved and the innovative features; c. Drawings and illustrations: the user uploads technical drawings and illustrations that support the understanding of the invention, d. Claims: the wizard includes templates and examples to help you formulate claims for protection of your invention,

[0048] And. Field of Application: the user specifies the potential fields of application of the invention, f. Technical antecedents: section dedicated to the description of the state of the art and related patents.

[0049] Advantageously, Al-based algorithms support the user by providing suggestions based on existing patents, ensuring that all necessary details are included.

[0050] Advantageously, the Al-based algorithms used in the present method are advanced to process patent data and generate visual content representing the invention and its detailed applications. Such content autogenerated by means of artificial intelligence algorithms also includes possible further uses of the product.

[0051] Specifically, the algorithms used are the following:

[0052] - Natural Language Processing (NLP): natural language processing algorithms extract key information from patent texts,

[0053] - Generative Adversarial Networks (GANs): used to generate images starting from textual descriptions, creating visual representations of the invention,

[0054] - Vision Transformers (ViT): algorithms that analyze images to identify specific components and features of the invention,

[0055] - Knowledge Graphs: used to identify and visually represent possible fields of application and correlations with other technologies. Similarly to what is described above, the fifth data processing module 50, for each patent, creates a visual output (images and / or videos) which represents the field of application, the context of possible use and the market trends and options sale and the relevant sales method is set (costs, type of licensing and discounts).

[0056] In accordance with the second module 20, search module illustrated in Figure 6, a specialized search engine using artificial intelligence means is able to perform a semantic search on the patent text and images and on all the elements created, the fifth data processing module 50 creates a visual output (images and / or videos) that represents the field of application, the context of possible use and the market trends and sales options.

[0057] To perform a search, simply type a short request and / or navigate the sections of the software platform by categories / technologies / images.

[0058] For example, you can use one of the five search methods, described below.

[0059] A first method is represented by the so-called "text-to-text search" by means of a portion of text, i.e. by inserting in the text search, for example, the wording "3D printing system", the software will provide a list of results based on of the semantic affinities found in the patent database.

[0060] A second method is represented by the so-called "text to image search" whereby by inserting in the text search, for example, the wording "curved light alloy screw" the software will provide a list of results based on the affinities found between the images recorded and created with the artificial intelligence means associated with the system, in the postprocessing phase.

[0061] A third method is represented by the so-called "image search" search by means of which by loading an image into the appropriate section, the software will provide a series of images similar to the one loaded. Furthermore, it will also be possible to search for a detail element within the image itself. For example, by loading an image of a car you can select the area with the wheel to search only for images similar to a wheel and if necessary you can ask the system to create a detailed image of a wheel with all the components to select the specific component of interest such as the bolt that secures the wheel.

[0062] A fourth method is represented by the so-called "use-case search", by means of which by inserting in the textual search, for example, the words "I need a metal valve of 2 cm in diameter with resistance specifications suitable for use in aerospace sector", the software will search for all patent documents for valves whose use involves use in the aerospace sector, but will also provide a list of auto-generated results (auto-generated Use- Cases) of possible valves patented for other fields but with properties equally suitable for use in the space sector, providing for example a result of the type: patent n.865390, entitled "valve in special alloy... for use in the sports sector", "Matching on self-generated Use-Cases": the alloy proposed in It also allows use in the aerospace sector.

[0063] A fifth method is represented by the so-called search by purpose, i.e. "patent collection search", in case you want to carry out a multiple use search on multiple patents of a plurality of components, to create a specific product starting from said components.

[0064] The "patents collection" function allows users to create a new product by combining multiple existing patents. This feature uses the power of artificial intelligence algorithms to break down a product into its fundamental components and search for the patents that best suit each component, taking into account the specific features requested by the user.

[0065] The "patent collection search", specifically, includes the following phases:

[0066] 1. Input on the Product:

[0067] -the user types the name and a brief description of the product he wishes to create. For example, "a green toaster with heat-resistant paint."

[0068] 2. Product Breakdown:

[0069] - by means of artificial intelligence algorithms, the description provided by the user is analyzed using advanced language models (LLM) to identify and break down the product into its fundamental components. For example, for a toaster, the components might include: outer shell, heating element, thermostat, paint, heating mechanism, etc.

[0070] 3. Search for Components:

[0071] - for each identified component, the Al performs a search in the patent database using a specific "query" that includes the features requested by the user. For example, if the user requests a heat-resistant and green paint, the system will search for patents that meet both features.

[0072] 4. Viewing Results:

[0073] - the search results are displayed in a list, showing for each component the patents found that satisfy the requested specifications. For example:

[0074] - external shell: patent for a resistant and lightweight metal alloy;

[0075] - resistance: patent for a high efficiency heating element.

[0076] - paint: patent for a heat-resistant green paint.

[0077] 5. Manually adding additional components:

[0078] - the user has the possibility to manually add components to the list if they have not been identified by the AL This allows for greater flexibility and product customization.

[0079] 6. Integration and Compatibility:

[0080] -Al provides suggestions on how to integrate the different patented components to assemble the final product, ensuring compatibility and performance optimization.

[0081] 7. Creation of the new product:

[0082] - once all the necessary patents have been selected, the user can view a detailed guide on how to combine the various components to create the new product. This guide includes assembly diagrams, technical specifications and practical tips.

[0083] A practical example is described below, in case a user wants to create a "green toaster" with the following characteristics:

[0084] - external shell: resistant and made of recyclable material.

[0085] - paint: green and heat resistant.

[0086] - resistance: high energy efficiency.

[0087] The phases of the process, as described above for the "patent collection search", are as follows: 1. information on the product to be made: "Green toaster with heat- resistant paint",

[0088] 2. decomposition: the Al breaks down the product into the following components: external shell, paint, resistance and all other secondary components,

[0089] 3. search: external shell: search for patents for recyclable and resistant materials; paint: search for patents for heat-resistant green paints; resistance: search for patents for high efficiency heating elements,

[0090] 4. visualization of the Results: external body: patent n.123456 for a recyclable metal alloy; paint: patent n.789012 for a heat-resistant green paint; resistance: patent n.345678 for a high efficiency heating element,

[0091] 5. manual addition: the user manually adds a component for the "digital control panel" not identified by Al,

[0092] 6. integration: Al suggests how to assemble the product ensuring that all components work correctly together,

[0093] 7. creation of the new product: the user receives a complete guide for assembling the green toaster.

[0094] Advantageously, the "patents collection" function allows you to maximize the use of patents, through the use of patents that often remain inactive. It therefore valorizes existing inventions.

[0095] Advantageously, the "patents collection" function allows the creation of innovative products. It facilitates the creation of new products protected by intellectual property by combining different already patented technologies.

[0096] Advantageously, the "patents collection" function allows great flexibility and customization. Users can add and modify components, tailoring the final product to their specific needs.

[0097] In the presence of patents that anticipate the invention described, found by the search carried out, the method object of the present invention activates a wizard to help the user compose a new patent text aimed at overcoming the patents found by this search. The wizard is similar to the one already described for drafting a patent text, but includes further correction / checking phases by Al-based algorithms:

[0098] - gap analysis: Al-based algorithms identify areas not covered by existing patents and suggest how to improve the description of the invention, -wording suggestions: Al-based algorithms provide examples of effective wordings for claims and technical description,

[0099] -similarities and differences: Al-based algorithms highlight similarities and differences compared to existing patents to help the user emphasize the uniqueness of their invention.

[0100] The third module 30, e-commerce module, allows interested users to use the patents under license, for example by paying an annual subscription to the platform itself. This subscription may be calibrated on the number of patents, by category and / or by characteristics deemed of interest. The license is based, for example, on the standard used by platforms for sharing and selling graphic or musical content. If of interest to users, the software platform also allows you to manage the purchase of an exclusive license to use a patent.

[0101] The systems and / or subscription plans may for example include: -monthly subscriptions for stocks and / or categories of patents - annual subscriptions for stocks and / or categories of patents

[0102] - one-off credits for stocks and / or categories of patents

[0103] - purchasing process for an exclusive patent license.

[0104] Conveniently, one-off credits can be acquired in the form of a "premium" subscription to the software platform.

[0105] For each granted patent, the system will display the cost and / or credits needed and the related license methods that can be purchased :

[0106] - Exclusive license: with an exclusive license, the patent owner grants a single party the right to use the invention exclusively. This means that the patent holder cannot grant other licenses for the invention to third parties and can only use it in a limited way, if any. This type of license gives the buyer a competitive advantage in the market.

[0107] - Non-exclusive license: With a non-exclusive license, the patent owner grants one or more parties the right to use the invention. This means that the patent holder will be able to grant multiple licenses for the same invention to third parties.

[0108] - Limited License: A limited license establishes restrictions on the use of the invention. For example, the patent holder may grant a limited license only for a specific industry or for a limited period of time. This type of license is useful when the patent owner wishes to maintain a certain degree of control over the use of the invention. - Cross licensing: A cross license is an agreement between two parties who own different patents and grant each other the right to use each other's inventions. This type of license is often used to encourage collaboration and encourage the development of new technologies.

[0109] -Sublicense: A sublicense is a license granted by a licensee to another party, allowing the latter to use the invention covered by the patent. This type of license is usually used when the original licensee wishes to further expand the distribution or use of the invention.

[0110] - Sale or Assignment: Another option is to sell or assign the patent to another interested party. This can be done through a direct purchase transaction or an assignment of patent rights. The sale or assignment of the patent may involve a single payment or a deferred payment, depending on the agreements reached.

[0111] - Royalties: If you grant licenses for use, you may receive royalties as compensation for using the patent. Royalties may be based on a percentage of sales of products or services using the patent or a fixed fee per unit sold.

[0112] The method according to the present invention also allows us to avoid commercially penalizing the patents considered furthest from the "core business" of global interest today, estimated at a percentage of approximately 65% of the patents filed. The technologies described in these patents could instead prove useful for widespread use and could also contribute to eliminating the problem of "Patent Trolls".

[0113] Another aspect of the present invention concerns predictive marketing with "intelligent" data. In fact, the system object of the present invention, by analyzing market trends, will carry out a predictive analysis on the future uses of the inventions for different application fields (use-cases). The platform will then be able to predict which patents (currently not for sale) will have the greatest probability of being sold in the following months.

[0114] In accordance with the fourth module 40, social module, the creation of user groups based on scientific and purchasing interests is envisaged. In particular through this fourth module 40, the method can recommend:

[0115] - groups of collaborators for those who want to enhance their IP portfolio;

[0116] - patents of its IP portfolio based on market trends and / or scientific interests.

[0117] The User Group Identification process includes the following steps:

[0118] - analyze publications: Al-based algorithms analyze scientific articles, publications and patents to identify experts in the relevant field.

[0119] - profiling users: based on declared interests and activity on the platform, Al-based algorithms suggest groups of collaborators with complementary skills,

[0120] - use market trends: use market data to identify emerging sectors and suggest strategic collaborations.

[0121] With reference to Figure 8, a general user can carry out the method according to the present invention by interfacing with the software platform. First you need to register and choose a strong authentication method. Once inside the software platform you can interact in the following way:

[0122] - importing patent data from a database in automatic or manual mode,

[0123] - obtaining the processing of said data, with the generation of images and videos of the same and a list of possible fields of application of the technology,

[0124] - searching for a patent through a text / image / field of application search and / or through the correlation of multiple patents for the creation of a specific product,

[0125] - in the event of a negative outcome of the search, using a wizard, composing a new patent text to be subsequently submitted to a patent office,

[0126] - transacting a commercial operation, for example, purchase / sale of a patent, acquisition / granting of a license,

[0127] - identifying groups of collaborators to enhance their IP portfolio and / or patents based on market trends and / or scientific interests.

[0128] In addition to the embodiment of the invention, as described above, it should be understood that numerous other variations exist. It should also be understood that such embodiments are exemplary only and do not limit either the scope of the invention, its applications, or its possible configurations. On the contrary, although the above description allows the specialized technician to implement the present invention at least according to one of its exemplary embodiments, it must be understood that many variations of the described components are possible, without thereby departing from the scope of the invention, as defined in the attached claims, which are interpreted literally and / or according to their legal equivalents.

Claims

1. Method for managing patents by means of artificial intelligence algorithms implemented on a computer system, the method comprising the following steps:- importing patent data from a database in an automatic or manual way, or if the patent does not exist by activating a wizard to draft a patent application,- processing the imported patent data, using artificial intelligence algorithms to generate images and videos of the same and a list of possible fields of application of the technology inherent to the patent data,- searching for a patent through a text / image / field of application search and / or self-generated "use-cases" and / or through the correlation of multiple patents to extract the characteristics of a specific product from these patents,- composing a new patent text to be subsequently submitted to a patent office, in the event of a negative outcome of the search, using a wizard,- transacting a commercial operation, for example, purchase / sale of a patent, acquisition / granting of a license,- identifying groups of collaborators to enhance their IP portfolio and / or patents based on market trends and / or scientific interests.

2. Method according to claim 1, wherein the automatic mode comprises the acquisition of patents through search indexes which extract the patents from existing databases and register the patents by classifying them on the basis of the structural data present.

3. Method according to claim 1, in which the manual mode of data entry includes the insertion of a specific patent and the creation of a series of possible modes of use alternative to the main one.

4. Method according to claim 3, in which the identification of the possible alternative methods of use includes the further phases:-analyzing the patent inserted with other similar inventions in order to identify uses that are not immediately obvious, using artificial intelligence algorithms,-simulating scenarios through simulation models to predict how the invention can adapt to different application contexts,-suggesting, through analysis of historical data, the use of similar technologies.

5. Method according to any of the previous claims, wherein a wizard for drawing up a draft patent application includes the following steps: a. entering the title of the invention; b. describing the invention including the technical problem solved and the innovative features; c. inserting technical drawings, d. formulating claims, e. specifying the potential fields of application of the invention, f. inserting the state of the art.

6. Method according to claim 5, including the further correction / control phases by the artificial intelligence algorithms:- identifying areas not covered by existing patents and improve the description of the invention,-suggesting examples of effective formulations and technical description,- highlighting similarities and differences compared to existing patents.

7. Method according to any of the previous claims, wherein the identification of user groups comprises the further steps:- analyzing publications using artificial intelligence algorithms,- profiling users based on declared interests and activity on the platform and suggesting groups of collaborators with complementary skills, using artificial intelligence algorithms,- using market trends to identify emerging sectors and suggest strategic collaborations.

8. Method according to any of the previous claims, wherein the search is carried out from text to text, whereby by inserting a portion of text into the textual search, the method provides a list of results based on the semantic affinities found in the patent database.

9. Method according to any of the previous claims, wherein the search is carried out from text to image, by means of which by inserting a portion of text into the textual search, the method provides a list of results based on the affinities found between the recorded images and created through artificial intelligence algorithms associated with the same method.

10. Method according to any of the previous claims, in which the search is carried out by image, by means of which by loading an image within the appropriate section, the method provides a series of images similar to the one loaded or a detail element to the inside the image itself.

11. Method according to any of the previous claims, in which the search is carried out by mode of use by which by inserting a specific modeof use into the textual search, the method will search for all patent documents whose use involves use in the field wanted.

12. Method according to any of the previous claims, wherein the search is carried out by collecting patents, by means of which a multiple search is carried out on multiple patents of a plurality of components, to create a complex product starting from the characteristics of the plurality of components.

13. Method according to claim 12, wherein the multiple patent search includes the further steps:- indicating the product you want to create;- breaking down the product to be created into its main components by means of artificial intelligence algorithms;- searching for patents relating to its main components;- viewing the patents found by the search,- manually adding any components not identified by the artificial intelligence algorithms,- suggesting using artificial intelligence algorithms how to assemble the new product, so that all components work correctly,- receiving a detailed guide on how to assemble the various components and obtaining the new product14. Method according to any of the previous claims, wherein the phase of transacting a commercial operation involves the acquisition of one- off credits for stocks and / or categories of patents.

15. Software platform for operating the method according to any of claims 1-14.

16. Computer system configured to perform the method according to any of claims 1-14.

Images