81 results about "Patent document" patented technology

The invention discloses a patent literature retrieval system, method and equipment and a storage medium, and relates to the field of data processing. The system comprises a data acquisition module for acquiring a retrieval request and historical behavior data of a user; the intention analysis module is used for analyzing the current request through a semantic analysis model and a technical field topic model so as to determine a user intention and an intention matching degree; the user behavior factor generation engine extracts dominant and recessive behavior characteristics from the historical data, and inputs the dominant and recessive behavior characteristics into a pre-trained user portrait model to generate user behavior factors; a recommendation generation engine recall candidate patents according to the intention of the user, and an authority calculation module calculates the authority of each patent according to a patent citation network; and the patent recommendation module fuses the user behavior factors, the user intention matching degree and the authority degree, and comprehensively sorts candidate patents to generate a recommended patent list. The method can meet the precision requirements of different users, and reduces the deviation between the patent recommendation result and the real expectation of the user.

Mutual authentication and CSR verification techniques are described in this patent document. When a first person calls a second person, neither of them know that the other person is who he or she says he or she is. After a second person receives the call, the second person can log into a provider portal using a user device. After the second person logs in, the second person can see a verification of the call, can input on the user device a time passcode, or can receive such as passcode from a central system to authenticate the first person. The first person can provide the passcode to the second person. Upon receiving the inputted passcode, the second person can use his or her user device to indicate that the time passcode is correct so that the second person can be authenticated to access the first person's account.

Systems and methods for reducing duplication of data in a data repository storing evidence corresponding to claim elements of a patent document. The systems and methods include a first user interface with a first plurality of rows corresponding to a plurality of first claim elements of a first patent document and a second plurality of columns corresponding to a plurality of second claim elements of a second patent document. The first plurality of rows and the first plurality of columns are arranged in a matrix format and form a plurality of cells. The user interface includes a user selectable input area for the plurality of cells. A pointer can be created in the data repository that links a first claim element of the plurality of first claim elements with a second claim element of the plurality of second claim elements responsive to an input in the user selectable input area.

The embodiment of the application relates to a processing method and device for mining patent information in combination with a large language model, the method comprising: selecting a large language model as a first model, assigning five target tasks to the first model, and constructing five task data sets to fine-tune the first model; receiving a PDF file of a small molecule drug patent after fine-tuning is completed and performing image, text block and table extraction on the PDF file; then performing molecular image / molecular label / irrelevant table screening and molecular activity table splicing based on the model; then constructing a label-image corresponding relationship table based on the molecular image / molecular label / spliced table information sequence and the first model, and constructing a label-activity information corresponding relationship table based on the first model and the spliced table information sequence; and merging the two corresponding relationship tables to obtain a multi-modal comprehensive information table; finally, a corresponding mining report is fed back by the patent file and the multi-modal comprehensive information table. Through the application, the data mining efficiency can be improved.

The specification discloses a patent retrieval method, device and equipment for multiple retrieval scenarios. Patent field language models are obtained by training general language models with fields of patent documents as training data. For multiple retrieval scenarios, positive and negative samples of each patent document are screened through different weight multi-path recall and random sampling, and training data of a patent semantic representation model of each retrieval scenario is constructed to train the patent field language models, and then the patent semantic representation model of each retrieval scenario is obtained to generate semantic vectors of each field of the patent, and the semantic vectors are stored in a patent retrieval vector database of each retrieval scenario. According to the semantic vectors of the retrieval fields, similar patent documents are searched from the patent retrieval vector database of the target retrieval scenario. Through the above scheme, the semantic representation ability of the patent semantic representation model for the patent fields of different retrieval scenarios is improved, and the retrieval precision is improved.

To provide a patent document creation device, etc. being further user-friendly in supporting creation of a patent document.SOLUTION: A patent document creation device for supporting creation of a patent document using a large-scale language model, includes an acquisition unit that acquires patent information relating to a patent to be described in the patent document, a question generation unit that generates a first question for requesting a detailed description of the patent using the large-scale language model on the basis of the patent information, an answer acquisition unit that acquires an answer to the first question from a user, an answer generation unit that when the answer to the first question is not acquired from the user, generates an answer to the first question using the large-scale language model, and a creation unit that creates the patent document using the large-scale language model on the basis of the patent information and a first user answer to the first question acquired from the user or a first automatic answer to the first question generated by the answer generation unit.SELECTED DRAWING: Figure 2

The invention provides a patent document database construction method and device based on technical description. The method comprises the following steps: identifying a main body tag and a description tag of each patent document based on a named entity identification model; carrying out punctuation mark-based feature statement division on the technical description part, and combining all main body tags and / or all description tags belonging to the same feature statement to obtain a technical description combination; and associating the patent number of the patent file with all the technical description combinations to form a patent file database based on the technical description. The patent document is subjected to entity recognition, the feature statements are used as the combination range of the main tags and the description tags, and the technical description of the patent document is combined by using more simplified main information, so that complete coverage of the feature information is realized; and the problem that the retrieval precision is influenced by redundant descriptions or interference words in traditional independent keywords or key sentences is avoided.

By presenting users with highly accurate patent maps, the system streamlines the process of reading patent documents. [Solution] The patent analysis system 10 performs the following steps: obtains a patent list from artificial intelligence 14 that includes the usage forms of substances extracted from patent documents; creates a patent map that visualizes the quantitative distribution of patent documents in the patent list using the usage forms and other indicators contained in the patent documents; and outputs the patent map to the user terminal 12.

The application provides a retrieval method and device, computer equipment and a storage medium, which can simultaneously consider the breadth and accuracy of patent document retrieval. The method comprises the following steps: receiving a search operation instruction input by a user, the search operation instruction being associated with attention data; determining a first patent data set according to the search operation instruction; performing retrieval grouping on the first patent data set according to a template file to obtain a plurality of tree data structure subgroups, the tree data structure subgroup comprising one or more tree data structure nodes, the template file being a tree data structure retrieval grouping file, each of the plurality of tree data structure subgroups corresponding to a group of tree data structure retrieval formulas in the template file; determining a patent data set corresponding to the tree data structure subgroup, the patent data set being a data set arranged according to a list data structure; and performing semantic analysis on the patent data set to obtain a plurality of semantic groups.

The application provides a multilingual patent translation dataset optimization system and method based on a large language model. The system includes a data processing module for obtaining an initial translation dataset of multilingual patent documents, performing quality evaluation on each original text-initial translation pair in the initial translation dataset to obtain a current quality score of each pair; a data filtering module for filtering out original text-initial translation pairs with a current quality score lower than a quality threshold as a to-be-optimized data block according to the quality threshold; an iterative repair optimization module for performing an iterative repair optimization loop on the to-be-optimized data block until its quality score reaches the quality threshold or meets an iteration termination condition; and an optimized dataset output module for outputting data blocks meeting the quality threshold and their corresponding optimized translations as an optimized translation dataset. The application supports multi-GPU parallel computing and concurrent API calling to significantly improve the efficiency and quality of patent document processing.

A method for classifying patent documents may include the steps of identifying original patent documents in a patent pool identified according to a search and / or selection input of a user; determining the first patent documents classified in the classification generated by the user, among the original patent documents in the patent pool; outputting an indication for accuracy related to the automatic classification; and performing the automatic classification for second patent document which are not yet classified, among the patent documents in the patent pool.

To support quick and precise reading of description of a scope of a patented claim.SOLUTION: The present invention is directed to an information processing apparatus having an analysis unit for analyzing text data indicating a scope of a patent claim included in patent literature data to specify a claim element of invention for each of claims included in the scope of the patent claim, and a display control unit for displaying, on a display device, the text indicating each claim of the scope of the patent claim with dividing the text into portions corresponding to the respective claim elements.SELECTED DRAWING: Figure 3

The present invention relates to an apparatus and a method for processing, analyzing, and visualizing a scientific and technological raw data file, which can standardize and unify multiple names for a country, multi-named companies, and multi-named universities in a process of processing a scientific and technological raw data file including patent documents or papers, thereby reducing distortion of a data analysis result and improving accuracy, wherein the scientific and technological raw data file is uploaded to a data processing module, the multiple names for a country, multi-named companies, and multi-named universities of the scientific and technological raw data file are unified and inputted in a new first data processing file, the first data processing file is uploaded to a data analysis module to perform data indexing and analysis, the result is inputted in a second data processing file, and the second data processing file is uploaded to a data visualization module to perform visualization.

The invention discloses a technology transfer opportunity prediction method and system based on science-technology knowledge hysteresis. The method comprises the following steps: uniformly mapping scientific papers and patent documents to subdivision technology tags by utilizing a large language model based on fine tuning of a standard classification system; respectively constructing a time sequence scientific knowledge network and a technical knowledge network; calculating a knowledge lag value between each pair of technologies by using a time-lag cross-correlation algorithm; constructing a double-flow graph attention network model, respectively learning a historical transfer structure and a scientific lag attribute through a structure flow and an attribute flow, and fusing features by using a gating mechanism; and finally, carrying out multi-step prediction by adopting a sliding window strategy. According to the method, the problems of cross-modal data semantic alignment, scientific pilot quantization and time sequence attribute fusion are effectively solved, and the prospective accurate prediction of the technology transfer opportunity is realized.

The invention provides a design method and device of a patent retrieval statement, a medium and equipment. The design method comprises the following steps: identifying a logical operation symbol in a retrieval formula; based on the logic operation symbols, splitting the retrieval formula into at least one retrieval word group; wherein the search word group comprises at least one search word; identifying the category of the search word group; based on the category of the search word group, generating an extended search word of the search word group; based on the category of the search word group, the search word group and the extended search word, generating a search type search statement; according to the method, a retrieval formula is split into retrieval word groups according to logical operation symbols, extended retrieval words are generated by identifying the categories of the retrieval word groups, and more accurate retrieval statements are generated by combining the categories of the retrieval word groups, the retrieval word groups and the extended retrieval words, so that patent documents more conforming to expectations are provided for retrieval personnel, and the retrieval efficiency is improved. The retrieval difficulty is reduced; and the retrieval efficiency and effect are improved.

To provide a linear power generator using centrifugal force.SOLUTION: A rotary gear B1 and a rotary gear B2 attached to a rotary gear A having a horizontal shaft as a rotary shaft are engaged with a fixed gear D by utilizing a propulsive force generation technique described in patent Document 1 and a technique of a power generation device using centrifugal forces, gravity, and magnetic forces described in patent Document 2. When a rotary shaft A is rotated, a B1 part of the rotary shaft and a movable weight magnet C attached to the B2 part of the rotary shaft and moving linearly are moved at a high speed in a solenoid for power generation by the action of centrifugal forces when the solenoid for power generation coincides with a straight line from the center of the rotary shaft A toward the outside, and an induced current is generated.SELECTED DRAWING: Figure 1

The present invention relates to a patent document information management system, comprising: a permission management module, which is used for acquiring an account and a password of a user, and determining an operation permission of the user account for each module of the patent information management system; a basic-information entry module, which is used for storing basic information of a patent application, wherein the basic information at least comprises a patent application title, an applicant, an inventor, a contact telephone number of the inventor, and an e-mail address of the inventor; a customer correspondence module, which is used for storing a patent disclosure document of the applicant; an official-document management module, which is used for storing a plurality of official documents issued and outputted by a state authority for each patent application; and a filing document management module, which is used for storing a plurality of documents filed with the state authority that correspond to each patent application submitted thereto.

The application provides a patent automatic classification method and device, a terminal and a storage medium, and comprises the following steps: performing dependency syntax analysis on a patent document to be classified to obtain pretreatment information; performing word embedding on the pretreatment information to obtain document representation; and obtaining a flow category prediction result, a function category prediction result and a meta-function category prediction result based on the document representation and trained flow classifiers, function classifiers and meta-function classifiers. Through dependency syntax analysis and word embedding on the patent document to be classified, the flow category prediction result, the function category prediction result and the meta-function category prediction result are obtained by using machine learning technology, the comprehensiveness of the classification of the patent document to be classified is realized, and the problem that the patent automatic classification method in the prior art is very limited and cannot comprehensively classify patents is solved.

Compared to the frame-type booth furniture described in Patent Document 1, the objective is to provide a unit system that can loosely define a space while distinguishing it from other spaces. [Solution] The unit system (1) comprises at least one space unit (10) having a three-dimensional frame and at least one deck unit (20) having a floor frame (21), wherein the space unit (10) and the deck unit (20) are adjacent to each other.

We provide technology that leverages large-scale language models in SDI to deliver new value. [Solution] The patent document analysis program causes the computer to perform the following processes: receiving input of pre-configured settings for the analysis of patent documents; obtaining the first newly arrived patent document from a patent database; inputting a first instruction sentence for analyzing the content of the first newly arrived patent document according to the pre-configured settings into a large-scale language model, and generating analysis results for the first newly arrived patent document based on the results obtained from the large-scale language model; and outputting the analysis results.

The invention provides a patent technology brief report intelligent pushing method and system based on a user portrait and a large model, and the method comprises the steps: collecting user dominant identity information and hidden interaction behavior data, and constructing a user portrait vector containing role labels and technical interest distribution based on time distance and interaction depth dynamic weighting; the method comprises the following steps: disassembling patent literatures to be pushed by utilizing a domain knowledge fine-tuning large language model, extracting triple and claim core features of'technical problems, technical means and technical effects', and mapping the triple and claim core features into patent value vectors; and calculating the semantic similarity of the two patents, and determining a target patent by combining technical novelty and competitive risk index weighted correction. And taking the triple as a generation substrate, calling a role attention mask mechanism, and generating and pushing a personalized technical brief report adaptive to a user role through differentiated recombination and rewriting of a large language model. And collecting user feedback data, and dynamically adjusting weighted parameters by using a reinforcement learning algorithm to form closed-loop optimization.

The invention provides a semantic retrieval method and device based on a patent document database, and the method comprises the steps: carrying out the preset processing of a retrieval text inputted by a user, so as to obtain a preset tag of the retrieval text; each preset label is correspondingly converted into a label word vector, each label word vector is used as a target word vector in sequence, m patent numbers with the highest similarity with the target word vector are determined in the patent file database, and scores of the patent numbers are determined according to the similarity between the patent numbers and the target word vector; constructing a score dictionary of each patent number, and determining a total score of each patent number; and feeding the k patent numbers with the highest total score back to the user as a retrieval result of the retrieval text. According to the patent document database retrieval method and device, the retrieval text is subjected to structural splitting, and in the retrieval process based on the patent document database, the retrieval result most meeting the user requirement is obtained by combining the score conditions of the retrieval result under different preset tags through finer-grained and more accurate retrieval information.

The server extracts a characteristic part of the disclosure described in the independent claim in the patent document. The server acquires the investigation result output from the generative model by inputting, to the generative model, an instruction text for instructing an investigation of the relevance between the characteristic part and the technical information. The server outputs the investigation result.

There is a problem in that it is not easy to generate suitable prompts to be applied to patent documents. [Solution] A program to be executed by a computer having a processor and a memory unit, wherein the processor executes the following steps: a task reception step in which task information related to a task to be performed on a patent document is received; a query reception step in which query information to be applied to the patent document is received; a document acquisition step in which patent documents related to patents are acquired; an extraction step in which predetermined extracted portions are extracted from the patent documents acquired in the document acquisition step according to the task received in the task reception step; and a generation step in which a prompt is generated that includes the query information received in the query reception step and the extracted portions extracted in the extraction step.

The present application belongs to the technical field of data processing, and particularly relates to an FTO risk analysis method and system based on a feature map. The method comprises: obtaining patent files and to-be-implemented scheme materials; performing constraint analysis on the claim text in the patent files to obtain constraint units with necessity levels; collecting associated evidence according to the constraint units to obtain a candidate evidence set comprising identification candidates and evidence source chains; determining feature influence values according to the uncertainty and necessity levels of the identification candidates; constructing a claim feature map based on the constraint units and the candidate evidence set, and constructing a to-be-implemented feature map based on the to-be-implemented scheme materials; performing constraint satisfaction calculation and consistency correction according to the evidence source chains to obtain a corrected satisfaction degree; and generating an FTO risk analysis result according to the corrected satisfaction degree and the feature influence values. The present application can reduce the misjudgment caused by local similarity and evidence splicing, and improve the explainability and traceability of the risk result.

This application relates to the field of patent analysis, and in particular to a computer-aided patent document clustering method. The method includes: receiving multiple sets of clustering seeds, each set containing a cluster name and at least one corresponding clustering attractor; using a processing unit to determine whether the key feature information of each patent document contains a certain clustering attractor (step A); executing step A whenever additional sets of clustering seeds are received, until the number of patent documents whose key feature information does not contain any clustering attractor is less than a certain threshold; associating patent documents whose key feature information contains a certain clustering attractor with the cluster name corresponding to that attractor, thereby obtaining the cluster name of each patent document. This application improves the accuracy and efficiency of patent clustering, and is a practical, accurate, and efficient patent clustering method that can help users obtain highly accurate clustering results with high efficiency.

The invention discloses a pre-training model bert-based patent classification method and system. The method comprises the steps of dividing patent documents into a plurality of technical scheme entity pairs through a pre-training model bert; and setting a keyword score calculation model, calculating the keyword score of each vocabulary in each technical scheme entity pair, performing similarity comparison on the vocabulary with the highest keyword score and the technical classification vocabulary in the IPC, and taking the technical classification vocabulary with the highest similarity in the IPC as the technical classification of the patent literature.

Extracorporeal membrane oxygenation systems and methods of use are disclosed. Several publications and patent documents are cited throughout the specification in order to describe the state of the art to which this invention pertains. Full citations of these references can be found throughout the specification. Each of these citations is incorporated herein by reference as though set forth in full.

Modem deep neural network (DNN) models have many layers with a single layer potentially involving large matrix multiplications. Such heavy calculation brings challenges to deploy such DNN models on a single edge device, which has relatively limited computation resources. Therefore, multiple and even heterogeneous edge devices may be required for applications with stringent latency requirements. Disclosed in the present patent documents are embodiments of a model scheduling framework that schedules multiple models on a heterogeneous platform. Two different approaches, model first scheduling (MFS) and hardware first scheduling (HFS), are presented to allocate a group of models for a service into corresponding heterogeneous edge devices, including CPU, VPU and GPU. Experimental results prove the effectiveness of the MFS and HFS methods for improving the inference speed of single and multiple AI-based services.

The invention discloses a patent data analysis method based on a dynamic context window, which belongs to the technical field of text data analysis, and comprises the following steps: sorting patent documents; decomposing each patent document into a plurality of semantic blocks, and outputting a block set after algorithm processing; selecting the block sets processed by the two algorithms, and calculating and storing the technical association probability between the two segments; soft clustering is carried out on all semantic blocks according to the technology association probability, and a technology theme membership degree vector is obtained; constructing retrieval numbers for the semantic blocks on the basis of technical topic membership degree vectors; establishing a quick retrieval tree; and according to the size of the context window of each semantic block, retrieving and expanding the context of the corresponding semantic block by adopting the quick retrieval tree, and finally outputting an expanded result. According to the patent data analysis method based on the dynamic context window, the problem that the analysis effect is poor due to the fact that an existing patent data analysis mode is limited by an upper window and a lower window is solved.