Associated patent determination method and device, equipment, medium and product

Abstract

The invention discloses an associated patent determination method and device, equipment, a medium and a product, and the method comprises the steps: determining a patent retrieval mode corresponding to a currently browsed patent when the currently browsed patent is detected; wherein the patent retrieval mode is semantic retrieval or non-semantic retrieval; and if the patent retrieval mode is semantic retrieval, determining a target associated patent corresponding to the currently browsed patent from the retrieval database based on the target content in the currently browsed patent and the retrieval text corresponding to the currently browsed patent. The problem that in the prior art, in the patent retrieval process, the retrieval result is not accurate only based on keyword matching or simple semantic analysis is solved, the patent better conforming to the retrieval intention of the user is rapidly and accurately determined according to the currently browsed patent of the user, and therefore the patent retrieval experience effect of the user is improved.

Description

Technical Field

[0001] This invention relates to the field of information retrieval technology, and in particular to a method, apparatus, device, medium, and product for determining related patents. Background Technology

[0002] Information retrieval is a key means for people to obtain effective information, assist in decision-making, and filter core content. Against this backdrop, how to accurately mine potential user needs from massive patent data and efficiently push related patents has become a key issue that urgently needs to be addressed in the field of intellectual property services.

[0003] Currently, patent recommendation methods mainly rely on keyword matching and simple semantic analysis, but these methods have significant limitations in practical applications. For example, when using keyword matching or simple semantic analysis for patent recommendation, the patent search results may not be accurate enough, and core related patents may be missed.

[0004] To address the aforementioned issues, the method for determining related patents needs to be improved. Summary of the Invention

[0005] This invention provides a method, apparatus, device, medium, and product for determining associated patents, in order to solve the problem of inaccurate search results during patent retrieval.

[0006] In a first aspect, embodiments of the present invention provide a method for determining associated patents, comprising: When a currently viewed patent is detected, the patent search method corresponding to the currently viewed patent is determined; wherein, the patent search method is semantic search or non-semantic search; If the patent search method is semantic search, then based on the target content in the currently viewed patent and the search text corresponding to the currently viewed patent, the target associated patent corresponding to the currently viewed patent is determined from the search database.

[0007] Secondly, embodiments of the present invention also provide an apparatus for determining associated patents, comprising: The retrieval method determination module is used to determine the patent retrieval method corresponding to the currently viewed patent when the currently viewed patent is detected; wherein, the patent retrieval method is semantic retrieval or non-semantic retrieval; The associated patent determination module is used to determine the target associated patent corresponding to the currently viewed patent from the retrieval database based on the target content in the currently viewed patent and the retrieval text corresponding to the currently viewed patent if the patent retrieval method is semantic retrieval.

[0008] Thirdly, embodiments of the present invention also provide an electronic device, comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for determining associated patents as described in any embodiment of the present invention.

[0009] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer instructions, which are used to cause a processor to execute the method for determining associated patents as described in any embodiment of the present invention.

[0010] Fifthly, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the method for determining associated patents as described in any of the embodiments of the present invention.

[0011] The technical solution of this invention determines the patent retrieval method corresponding to the currently viewed patent upon detection. The patent retrieval method can be semantic or non-semantic. If the patent retrieval method is semantic, then based on the target content in the currently viewed patent and the corresponding search text, the target related patent is determined from the search database. In this technical solution, the core technical features of the currently viewed patent can be determined based on its target content (claim text and / or specification text). Based on this, and combined with the patent retrieval method used by the user, a deeper analysis of the core technical features of the currently viewed patent, along with the search intent reflected in the user's input search text, allows for more precise selection of at least one target related patent with strong relevance to the currently viewed patent from the search database. This effectively improves patent retrieval efficiency and the accuracy of search results. This solves the problem of inaccurate search results in existing technologies that rely solely on keyword matching or simple semantic analysis during patent retrieval. It achieves rapid and accurate determination of patents that better match the user's search intent based on the user's currently viewed patent, thereby improving the user's patent retrieval experience. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the content of the embodiments of the present invention and these drawings without creative effort.

[0013] Figure 1 This is a flowchart of a method for determining associated patents according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of a method for determining associated patents according to Embodiment 2 of the present invention; Figure 3 This is a flowchart of a method for determining associated patents according to Embodiment 2 of the present invention; Figure 4 This is a schematic diagram of the structure of a device for determining related patents according to Embodiment 3 of the present invention; Figure 5 This is a schematic diagram of the structure of an electronic device that implements the method for determining the associated patent in the embodiments of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention. The acquisition, transmission, storage, use, and processing of data in the technical solutions of this application comply with the relevant provisions of national laws and regulations. It should be noted that in the embodiments of this application, certain software, components, or models and other existing solutions in the industry may be mentioned. These should be considered as exemplary, and their purpose is only to illustrate the feasibility of implementing the technical solutions of this application, but it does not mean that the applicant has or necessarily used such solutions.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in sequences other than those illustrated or described herein.

[0016] Example 1 Figure 1 The flowchart of a method for determining associated patents provided in Embodiment 1 of the present invention is applicable to situations where, based on the user's patent search method and the core content of the currently viewed patents, patents that better match the user's search intent are selected. This method can be executed by an associated patent determination device, which can be implemented in hardware and / or software and can be configured in a computing device capable of executing the associated patent determination method.

[0017] like Figure 1 As shown, the method includes: S110. When the currently viewed patent is detected, determine the patent search method corresponding to the currently viewed patent.

[0018] Here, "currently viewed patent" refers to the patent currently being viewed on the user's current browsing page. In this technical solution, the patent search method is either semantic search or non-semantic search.

[0019] Specifically, when users conduct patent searches on the patent system, they can enter keywords, patent summaries, technical issues, or patent requirement descriptions on the patent search page to find patents that meet their expectations. During this process, the patent system can recommend other patents with high relevance to the currently viewed patent, thereby helping users quickly find satisfactory patents.

[0020] It is understandable that when pushing relevant patents to users solely through keyword matching during the patent search process, the recommendation results may be inaccurate. However, semantic analysis of the search text entered by the user can effectively improve the accuracy of patent recommendations.

[0021] The reason is that the search text entered by the user usually includes more precise patent search requirements than keywords. Therefore, when the current patent being viewed by the user is detected in this technical solution, it is first determined whether the patent search method used by the user to search for the patent is semantic search. Optionally, the patent search method corresponding to the currently viewed patent is determined, including: retrieving the operation log corresponding to the currently viewed patent, and determining the patent search method of the currently viewed patent based on the operation log.

[0022] For example, the operation log of the user's search for patent A (i.e., the currently viewed patent) is retrieved. If the operation log shows that the user entered "image recognition patent based on deep learning" and selected the semantic search mode, then the patent search method is determined to be semantic search; if the user only entered the keyword "image recognition patent" and did not select the semantic search mode, then it is determined to be non-semantic search.

[0023] S120. If the patent search method is semantic search, then based on the target content in the currently viewed patent and the search text corresponding to the currently viewed patent, the target associated patent corresponding to the currently viewed patent is determined from the search database.

[0024] In this context, "target content" refers to information that characterizes the core content of the currently viewed patent. For example, target content includes the claims and / or specification of the currently viewed patent. "Search text" can be understood as the search query entered by the user when searching for a patent. "Search database" refers to the database of a patent search website or platform. "Target related patents" can be understood as one or more similar patents with a high degree of relevance to the currently viewed patent.

[0025] Based on the above example, if the user uses semantic search during the search, the search text "image recognition patent based on deep learning" is extracted, and patent matching is performed from the search database according to the semantic information represented by the search text. One or more similar patents with a high degree of relevance to the currently viewed patent are selected as target related patents.

[0026] When matching patents in the search database based on the search text, vector matching can be used. For example, a first vector corresponding to the search text and a second vector corresponding to the patents in the search database can be constructed. The vector similarity between the first vector and each second vector is calculated, and at least one target associated patent is determined based on the vector similarity. Optionally, the second vector corresponding to the patents in the search database can be pre-generated or generated in real time when the first vector is detected.

[0027] The advantage of this setup is that, while performing semantic patent retrieval, it matches target related patents based on the target content of the currently viewed patent (claims and / or specification) and the search text entered by the user. This not only accurately anchors the substantive content of the patent's technical solution itself, but also aligns with the user's search intent, achieving a more comprehensive, accurate, and practically relevant similar patent location than a single text search.

[0028] The technical solution of this invention determines the patent retrieval method corresponding to the currently viewed patent upon detection. The patent retrieval method can be semantic or non-semantic. If the patent retrieval method is semantic, then based on the target content in the currently viewed patent and the corresponding search text, the target related patent is determined from the search database. In this technical solution, the core technical features of the currently viewed patent can be determined based on its target content (claim text and / or specification text). Based on this, and combined with the patent retrieval method used by the user, a deeper analysis of the core technical features of the currently viewed patent, along with the search intent reflected in the user's input search text, allows for more precise selection of at least one target related patent with strong relevance to the currently viewed patent from the search database. This effectively improves patent retrieval efficiency and the accuracy of search results. This solves the problem of inaccurate search results in existing technologies that rely solely on keyword matching or simple semantic analysis during patent retrieval. It achieves rapid and accurate determination of patents that better match the user's search intent based on the user's currently viewed patent, thereby improving the user's patent retrieval experience.

[0029] Example 2 Figure 2 The flowchart illustrates a method for determining associated patents according to Embodiment 2 of the present invention. Optionally, the method for determining the target associated patents corresponding to the currently viewed patent from the retrieval database based on the target content in the currently viewed patent and the search text corresponding to the currently viewed patent is further refined.

[0030] like Figure 2 As shown, the method includes: S210. When the currently viewed patent is detected, determine the patent search method corresponding to the currently viewed patent.

[0031] S220. If the patent search method is semantic search, then a technical solution vector and / or semantic search vector are generated based on the target content of the currently viewed patent, and the target vector corresponding to the currently viewed patent is determined based on the technical solution vector and / or semantic search vector.

[0032] The target vector refers to the vector obtained by combining the technical solution vector and the semantic retrieval vector.

[0033] For example, different weights are assigned to the technical solution vector and the semantic retrieval vector, and a weighted average is performed based on the technical solution vector, the semantic retrieval vector, and their respective weights to obtain the target vector corresponding to the currently viewed patent.

[0034] Optionally, the target content includes the text of the claims and / or the text of the specification. Based on the target content of the currently viewed patent, a technical solution vector and / or a semantic retrieval vector are generated, including: extracting the text of the claims and / or the text of the specification from the currently viewed patent, and performing vector encoding on the text of the claims and / or the text of the specification to obtain the technical solution vector corresponding to the currently viewed patent; and / or performing vector encoding on the retrieval text corresponding to the currently viewed patent to obtain the semantic retrieval vector corresponding to the currently viewed patent.

[0035] In this technical solution, vector encoding is taken as an example using the Bidirectional Encoder Representations from Transformers (BERT) model based on Transformers.

[0036] In a specific example, such as Figure 2 As shown, when users conduct patent searches on the patent system, they can find patents that meet their expectations by entering keywords, patent summaries, technical issues, or patent requirement descriptions on the patent search page. At this point, the system determines the user's search method (i.e., patent search method) based on the operation log. If the patent search method is semantic search, the system extracts the user-input search text "image recognition patent based on deep learning," and uses the BERT model to semantically encode the search text, generating a semantic search vector. For non-semantic searches, no semantic vector needs to be generated. Simultaneously, the system uses the BERT model to vectorize the claims text and / or specification text of the currently viewed patent to obtain a technical solution vector.

[0037] Based on the above example, the target vector corresponding to the currently viewed patent is determined based on the technical solution vector and / or semantic retrieval vector, including: determining the semantic similarity between the search text and the currently viewed patent based on the first vector similarity of the semantic retrieval vector and the technical solution vector; determining the first weight corresponding to the semantic retrieval vector based on the text length and semantic similarity of the search text; determining the second weight corresponding to the technical solution vector based on the first weight; and determining the target vector corresponding to the currently viewed patent by performing weighted processing based on the semantic retrieval vector, the first weight, the technical solution vector and the second weight.

[0038] The technical solution vector can be a vector generated solely based on the text of the claims, a vector generated solely based on the text of the specification, or a vector generated by combining the claims and the specification.

[0039] For example, a claim vector corresponding to the claim text and a specification vector corresponding to the specification text are generated respectively. If the claim vector and the semantic retrieval vector have a high vector similarity, the claim vector is used as the technical solution vector of the currently viewed patent. If the specification vector and the semantic retrieval vector have a high vector similarity, the specification vector is used as the technical solution vector of the currently viewed patent. Alternatively, the technical solution vector of the currently viewed patent can also be a vector obtained by weighting the claim vector and the specification vector.

[0040] Based on this, the first vector similarity is calculated according to the cosine similarity between the semantic retrieval vector and the technical solution vector, and the semantic similarity between the retrieved text and the currently viewed patent is calculated through semantic analysis. Furthermore, the first weight corresponding to the semantic retrieval vector is determined based on the text length and semantic similarity of the retrieved text.

[0041] The advantage of this setup is that the longer the search text is, the more precise the user's search direction is, and the more attention should be paid to content related to the search text. Correspondingly, the higher the similarity between the search text and the currently viewed patent text, the more the currently viewed patent matches the user's search intent. Therefore, the longer the text length and the higher the text similarity, the higher the first weight needs to be assigned to the technical solution vector, thereby increasing the likelihood of searching the patent search database for the patent most similar to the currently viewed patent and making patent recommendations to the user.

[0042] In a specific example, see [link to example]. Figure 2 The first weight of the technical solution vector can be determined using the following formula: in, Indicates the first weight. α The basic allocation coefficient represents the text length and text similarity of the text to be retrieved. f(L) The normalized contribution function represents the text length of the text to be retrieved. The function representing the correction contribution of the similarity between the semantic retrieval vector and the corresponding part of the patent is used. The interaction enhancement factor represents the text length and text similarity of the text to be retrieved.

[0043] in, This is used to reflect the extra weight given to long sentences with high similarity, avoiding bias caused by a single factor.

[0044] in, , A quantized value representing the length of the text to be retrieved. This can be expressed by the following formula: Among them, and represent the 25th percentile and the 75th percentile, representing the short text threshold and the long text threshold respectively, and can also be set by yourself.

[0045] Among them, represents the number of valid words in the retrieved text (e.g., the number of content words after filtering stop words, such as "of, and" etc. are not included), is the confidence correction for text length. If the retrieved text contains strong limiting words such as "must, only, for" etc. (matched through the domain word list), then , otherwise it is 1.

[0046] Among them, , is the cosine similarity between the semantic retrieval vector and the technical solution vector (that is, the text similarity between the retrieved text and the text of the currently viewed patent).

[0047] Among them, can be expressed by the following formula: Among them, is the exceeding ratio of long sentences, represents the exceeding value of the maximum text similarity. The product of the two ensures that the longer the text length of the retrieved text, the greater the text similarity, then is greater.

[0048] For example, count the number of valid words L of the retrieved text "Patent for Image Recognition Based on Deep Learning". After filtering the stop word "of", the number of valid words L = 6. Since , , and 5 < 6 < 15, calculate according to the text length quantization value formula of the retrieved text. Since the retrieved text contains the strong limiting word "based on", the length confidence correction C_L = 1.1. Given that the first vector similarity S between the semantic retrieval vector and the technical solution vector is 0.85, calculate according to the g(S) calculation formula. Since L = 6 < Q_75 = 15, h(L, S) = 1.0. Substitute it into the formula of the first weight W1 to obtain the first weight corresponding to the semantic retrieval vector. On this basis, the second weight of the technical solution vector can be expressed as W2 = 1 - W1.

[0049] Furthermore, perform weighted processing on the semantic retrieval vector, the first weight, the technical solution vector and the second weight to determine the final semantic vector (that is, the target vector) of the currently viewed patent. For example, the target vector can be expressed as: Target vector = W1 × Semantic retrieval vector + W2 × Technical solution vector.

[0050] The weighting method depends on the text length and text similarity of the text to be retrieved, as shown above.

[0051] S230. For at least one candidate patent in the search database, based on the vector similarity between the candidate vector and the target vector corresponding to each candidate patent, determine the target associated patent corresponding to the currently viewed patent from at least one candidate patent.

[0052] Candidate patents can be any patent document in the search database, or patent documents selected from the search database based on keywords in the currently viewed patent or the text to be searched. A candidate vector is a vector obtained by vector encoding the claims and / or specification text of a candidate patent based on the encoding method corresponding to the text to be searched.

[0053] For example, based on the vector similarity between the candidate vector and the target vector corresponding to each candidate patent, the target associated patent corresponding to the currently viewed patent is determined from at least one candidate patent, including: calculating the second vector similarity between the candidate vector and the target vector of at least one candidate patent respectively, and sorting the corresponding candidate patents from high to low based on each second vector similarity; and determining the target associated patent corresponding to the currently viewed patent from at least one candidate patent based on the sorting result.

[0054] Specifically, the candidate patents in the search database are encoded with the same vector to obtain the candidate vectors corresponding to the candidate patents. The similarity between the candidate patents and the currently viewed patents is determined based on the cosine similarity between the candidate vectors and the target vectors. This allows for the identification of one or more target related patents that are most closely associated with the currently viewed patents from at least one candidate patent (i.e., outputting recommendation results).

[0055] The advantage of this setup is that it allows for the quantitative ranking of candidate patents based on the similarity of their semantic vectors, prioritizing the selection of patents that are closest to the currently viewed patent technology content. This improves the accuracy, objectivity, and retrieval efficiency of identifying target-related patents and avoids biases caused by subjective judgment.

[0056] The technical solution of this invention, upon detecting a currently viewed patent, determines the patent retrieval method corresponding to that patent. If the patent retrieval method is semantic retrieval, a technical solution vector and / or semantic retrieval vector are generated based on the target content of the currently viewed patent, and a target vector corresponding to the currently viewed patent is determined based on the technical solution vector and / or semantic retrieval vector. For at least one candidate patent in the retrieval database, a target associated patent corresponding to the currently viewed patent is determined from at least one candidate patent based on the vector similarity between the candidate vector and the target vector corresponding to each candidate patent. In this technical solution, the core technical features of the currently viewed patent can be determined based on the target content (claim text and / or specification text). On this basis, combined with the patent retrieval method used by the user, based on a deep analysis of the core technical features of the currently viewed patent, and according to the retrieval intent reflected in the user's input text, at least one target associated patent with a strong correlation to the currently viewed patent can be more accurately selected from the retrieval database, effectively improving patent retrieval efficiency and the accuracy of retrieval results. This technology solves the problem of inaccurate search results when relying solely on keyword matching or simple semantic analysis in the patent search process. It enables the rapid and accurate identification of patents that better match the user's search intent based on the patents currently being viewed, thereby improving the user's patent search experience.

[0057] Example 3 Figure 4 This is a schematic diagram of a device for determining related patents provided in Embodiment 3 of the present invention. Figure 4 As shown, the device includes a retrieval method determination module 310 and an associated patent determination module 320.

[0058] The retrieval method determination module 310 is used to determine the patent retrieval method corresponding to the currently viewed patent when the currently viewed patent is detected; wherein the patent retrieval method is semantic retrieval or non-semantic retrieval. The associated patent determination module 320 is used to determine the target associated patent corresponding to the currently viewed patent from the retrieval database based on the target content in the currently viewed patent and the retrieval text corresponding to the currently viewed patent if the patent retrieval method is semantic retrieval.

[0059] The technical solution of this invention determines the patent retrieval method corresponding to the currently viewed patent upon detection. The patent retrieval method can be semantic or non-semantic. If the patent retrieval method is semantic, then based on the target content in the currently viewed patent and the corresponding search text, the target related patent is determined from the search database. In this technical solution, the core technical features of the currently viewed patent can be determined based on its target content (claim text and / or specification text). Based on this, and combined with the patent retrieval method used by the user, a deeper analysis of the core technical features of the currently viewed patent, along with the search intent reflected in the user's input search text, allows for more precise selection of at least one target related patent with strong relevance to the currently viewed patent from the search database. This effectively improves patent retrieval efficiency and the accuracy of search results. This solves the problem of inaccurate search results in existing technologies that rely solely on keyword matching or simple semantic analysis during patent retrieval. It achieves rapid and accurate determination of patents that better match the user's search intent based on the user's currently viewed patent, thereby improving the user's patent retrieval experience.

[0060] Optionally, the associated patent determination module includes: a target vector determination unit, used to generate a technical solution vector and / or semantic retrieval vector based on the target content of the currently viewed patent, and to determine the target vector corresponding to the currently viewed patent based on the technical solution vector and / or semantic retrieval vector; The associated patent determination unit is used to determine the target associated patent corresponding to the currently viewed patent from at least one candidate patent in the retrieval database based on the vector similarity between the candidate vector and the target vector corresponding to each candidate patent.

[0061] Optionally, the target vector determination unit includes: a technical solution vector determination subunit, used to extract the claim text and / or specification text from the currently viewed patent, and to perform vector encoding on the claim text and / or specification text to obtain the technical solution vector corresponding to the currently viewed patent; and / or The semantic retrieval vector determination subunit is used to perform vector encoding on the retrieval text corresponding to the currently viewed patent to obtain the semantic retrieval vector corresponding to the currently viewed patent.

[0062] Optionally, the target vector determination unit includes: a semantic similarity determination subunit, used to determine the semantic similarity between the retrieved text and the currently viewed patent based on the first vector similarity between the semantic retrieval vector and the technical solution vector; The first weight determination subunit is used to determine the first weight corresponding to the semantic retrieval vector based on the text length and semantic similarity of the retrieved text. The second weight determination subunit is used to determine the second weight corresponding to the technical solution vector based on the first weight. The target vector determination subunit is used to determine the target vector corresponding to the currently viewed patent by performing weighted processing based on the semantic retrieval vector, the first weight, the technical solution vector, and the second weight.

[0063] Optionally, the associated patent determination unit includes: a sorting subunit, used to calculate the second vector similarity between the candidate vector and the target vector of at least one candidate patent, and sort the corresponding candidate patents from high to low based on each second vector similarity; The associated patent determination subunit is used to determine the target associated patent corresponding to the currently viewed patent from at least one candidate patent based on the sorting results.

[0064] Optionally, the device for determining associated patents is used to retrieve the operation log corresponding to the currently viewed patent and determine the patent search method for the currently viewed patent based on the operation log.

[0065] The associated patent determination device provided in the embodiments of the present invention can execute the associated patent determination method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0066] Example 4 Figure 5 A schematic diagram of the structure of an electronic device 10 according to an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0067] like Figure 5As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0068] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0069] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the method for determining the associated patent.

[0070] In some embodiments, the method for determining associated patents may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the method for determining associated patents described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the method for determining associated patents by any other suitable means (e.g., by means of firmware).

[0071] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0072] Computer programs used to implement the method for determining the associated patent of this invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0073] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0074] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0075] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0076] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0077] Example 5 This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the method for determining associated patents as provided in any embodiment of this application.

[0078] In implementing the computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0079] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0080] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for determining related patents, characterized in that, include: When a currently viewed patent is detected, the patent search method corresponding to the currently viewed patent is determined; wherein, the patent search method is semantic search or non-semantic search; If the patent search method is semantic search, then based on the target content in the currently viewed patent and the search text corresponding to the currently viewed patent, the target associated patent corresponding to the currently viewed patent is determined from the search database.

2. The method according to claim 1, characterized in that, The step of determining the target associated patent corresponding to the currently viewed patent from the search database based on the target content in the currently viewed patent and the search text corresponding to the currently viewed patent includes: A technical solution vector and / or semantic retrieval vector are generated based on the target content of the currently viewed patent, and a target vector corresponding to the currently viewed patent is determined based on the technical solution vector and / or the semantic retrieval vector. For at least one candidate patent in the search database, based on the vector similarity between the candidate vector corresponding to each candidate patent and the target vector, a target associated patent corresponding to the currently viewed patent is determined from the at least one candidate patent.

3. The method according to claim 2, characterized in that, The target content includes the text of the claims and / or the text of the specification. The generation of technical solution vectors and / or semantic retrieval vectors based on the target content of the currently viewed patent includes: Extract the claims text and / or specification text from the currently viewed patent, and perform vector encoding on the claims text and / or specification text to obtain the technical solution vector corresponding to the currently viewed patent; and / or The search text corresponding to the currently viewed patent is vector-encoded to obtain the semantic search vector corresponding to the currently viewed patent.

4. The method according to claim 2, characterized in that, Determining the target vector corresponding to the currently viewed patent based on the technical solution vector and / or the semantic retrieval vector includes: Based on the first vector similarity between the semantic retrieval vector and the technical solution vector, the semantic similarity between the retrieved text and the currently viewed patent is determined; Based on the text length of the retrieved text and the semantic similarity, a first weight corresponding to the semantic retrieval vector is determined; The second weight corresponding to the technical solution vector is determined based on the first weight; The target vector corresponding to the currently viewed patent is determined by weighting the semantic retrieval vector, the first weight, the technical solution vector, and the second weight.

5. The method according to claim 2, characterized in that, The step of determining the target associated patent corresponding to the currently viewed patent from at least one candidate patent based on the vector similarity between the candidate vectors corresponding to each candidate patent and the target vector includes: Calculate the second vector similarity between the candidate vector of at least one candidate patent and the target vector, and sort the corresponding candidate patents from high to low based on the second vector similarity. Based on the sorting results, the target associated patent corresponding to the currently viewed patent is determined from at least one candidate patent.

6. The method according to claim 1, characterized in that, Determining the patent search method corresponding to the currently viewed patent includes: Retrieve the operation log corresponding to the currently viewed patent, and determine the patent search method for the currently viewed patent based on the operation log.

7. A device for determining related patents, characterized in that, include: The retrieval method determination module is used to determine the patent retrieval method corresponding to the currently viewed patent when the currently viewed patent is detected; wherein, the patent retrieval method is semantic retrieval or non-semantic retrieval; The associated patent determination module is used to determine the target associated patent corresponding to the currently viewed patent from the retrieval database based on the target content in the currently viewed patent and the retrieval text corresponding to the currently viewed patent if the patent retrieval method is semantic retrieval.

8. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, which enables the at least one processor to perform the method for determining the associated patent according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for determining the associated patent as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method for determining the associated patent as described in any one of claims 1-7.

Images