Strategic Patent Search Automation

Method for searching for relevant patent documents using machine learning and AI techniques to improve the effectiveness and efficiency of patent searching. The method involves parsing claim elements, selecting keywords, identifying synonyms, forming search strings, ranking documents, determining citations, and mapping elements. This allows targeted searching of databases using user-defined elements instead of keywords, improving recall and reducing false positives compared to keyword-based searching.

Source

Patent search method and platform that improves efficiency and expands the range of retrieved patents by leveraging product and effect trees. When searching for patents using brief effects, the platform extracts hyponym words from the product and effect trees to find related patents beyond just matching the exact words. This provides a hierarchy of related concepts that expands the search beyond just the exact input words. The platform also suggests higher level concepts to broaden searches when exact words aren't found.

Source

A hybrid AI system for patent claims analysis that combines machine learning and user input to accurately search and analyze patent documents. The system allows users to initiate searches based on reference documents, narrow the search focus, and select sections. It uses ML to transform the document and claims into embeddings representing their meaning. Element-by-element searches find matches based on these embeddings. Confidence scores rank the results. The system provides feedback to improve future searches.

Source

More accurate patent searching and analysis using AI, specifically natural language understanding (NLU) and machine learning (ML) techniques, to improve search results and avoid missing relevant patents. The method involves analyzing patent claims and relevant text from a reference document using AI systems to determine the meaning of claim elements and search for matches to rank patents based on how well their claims match the reference document.

Source

A method to assist users in patent searching by enabling them to select and search within patent documents based on key criteria like classification codes and related tags. The system retrieves patent documents matching the initial search criteria. Users can then select one or more patents of interest and extract additional criteria like classification codes from them. The system generates a refined search using the extracted criteria to find more relevant patents. This iterative process allows users to navigate and refine patent searches using criteria specific to patents, enhancing search accuracy and relevance.

Source

Patent document search using diffusion models to improve accuracy and efficiency of patent retrieval. The method involves using diffusion models to generate keyword-oriented abstracts, claims, and sentences from a patent document. These expanded versions are then searched separately to find similar patents. The results are integrated and weighted to identify the most relevant patents. This expansion and integration process allows more comprehensive and accurate retrieval compared to just matching keywords.

Source

Intelligent system for fast and accurate patent search using natural language queries. The system takes user input of patent-related information like product names, self-made words, and company names. It then filters and displays relevant patents based on this input. The system leverages databases to find matching products, companies, and keywords. This improves the accuracy and efficiency of patent searching compared to querying raw patent databases.

Source

Automatically extracting patent document templates from a patent corpus by identifying sets of similar patent documents and extracting the common text shared between them. This allows finding and generating standardized document templates from a collection of similar documents like patents.

Source

A method for searching for similar patents based on natural language expressing most of the contents of a patent. The method involves extracting patent elements like components and functional elements from a query patent, extracting search words from the elements, matching and sorting elements of the query patent and elements of similar patents, calculating a matching rate, identifying missing elements, and optionally prompting a user to input a paraphrase for those missing elements. The paraphrase is then used to search for additional similar patents. This allows more accurate matching of components and handling of paraphrases and neologisms in patent descriptions.

Source

Automatically generate concise summaries of patent documents by extracting key information from the technical problem, solution, and beneficial effects sections of the patent. The method involves using pre-trained models to extract problem, solution, and effect sentences from the patent document. These sentences are then combined to create a summary of the invention.

Source

A method for creating a patent document summary that involves leveraging natural language processing and machine learning techniques to analyze the text of a patent document and generate a concise summary that captures key information from the document. The method involves identifying repetitive word sequences in the document, replacing subsequent occurrences of those sequences with shorter substitutes, and generating the summary based on the modified text.

Source

Automated prior patent search solution to quickly and easily generate patent search reports by leveraging text mining techniques. The solution involves extracting keywords from a search query, expanding them with synonyms, creating a search expression, searching patents, analyzing results, and generating a report. The text mining involves stopping unnecessary words, extracting primary keywords, expanding with synonyms, and storing unmatched keywords.

Source

Automated patent analysis and document generation system that can analyze patent claims and generate reports and search results based on the claims. The system uses natural language processing techniques to analyze patent claims and extract key elements. It generates search strings from these elements and sends them to external search engines to find related documents. The system also generates profiles for claims and documents based on linguistic analysis, syntactic categories, and metrics like stemming and word count. These profiles can be displayed to users for analysis.

Source

A platform for conducting patent searches and analyzing patent value using a combination of AI, machine learning, and blockchain technology. The platform leverages natural language processing to improve patent search relevance by analyzing similarities between claim limitations and prior art specifications. It also uses AI algorithms to rank prior art based on relevance to specific claim elements. The platform further provides tools for highlighting, weighting, and customizing search parameters.

Source

Patent information processing system using word frequency analysis to provide efficient patent searching and licensing recommendations. The system extracts key words from patents, calculates word frequencies, and uses similarity metrics to identify related patents. It also leverages linear programming to estimate patent licensing value based on factors like enterprise capabilities and patent descriptions. The system provides intelligent patent retrieval and suggests high correlation patents. It can also evaluate patent licensing potential through linear programming.

Source

Automatically classifying patents via machine learning to save time and resources. The method involves learning from a patent database to establish a basic classification model. When a user searches for patents, it uses their classification input to predict their personalized classification standard. This combines with the basic model to classify the remaining unclassified patents. It leverages transfer learning to customize the basic model using the user's classification pattern.

Source

Method and system for retrieving key technologies in patents and documents that improves upon existing patent retrieval methods by finding comprehensive and accurate key technologies in a field. The method involves extracting topics from patent and document databases, judging patent and document topic strengths, and identifying key techniques based on those scores. This analyzes documents and patents simultaneously to determine field-specific key technologies more comprehensively and accurately compared to just patent retrieval.

Source

Searching, filtering, and analyzing large numbers of patent-related documents to find documents of interest to a user. It provides user interface tools for filtering and analyzing patent-related documents to find documents of interest. It also allows searching for specific patent drawings.

Source

A method for accurately and efficiently finding similar patents to a target patent by analyzing the target claim and displaying candidate documents in order of similarity. The method involves extracting keywords from the target claim, finding candidate documents with those keywords, comparing the target claim to the candidates to determine similarity, and presenting the candidates in order of similarity. This allows focusing on the most relevant documents first and avoiding the need to scroll through long lists. The candidates are also displayed with highlighting to make the comparison easier.

Source

Patent search method and system that improves patent retrieval by generating more accurate and relevant search queries and expanding them to cover related patents. The method uses patent classification information and word embedding techniques to generate aspect queries and expand queries with double consistency. It also fuses search result sets using compound domain weights. This allows more effective retrieval by considering technical perspectives and semantic relatedness.

Source

Optimizing patent literature searches by taking a user's search formula and expanding it to a higher-quality search by adding relevant search terms and operators. The system receives a user's search formula, classifies it into groups based on operators, adds additional relevant search terms to each group, combines the groups with operators to create a final optimized search formula, and provides that to the user. This expands and improves the user's initial search.

Source

Semi-automatic patent infringement analysis using hybrid artificial intelligence. The system narrows the focus of patent search results by identifying relevant portions of a reference document. It then uses artificial intelligence to compare claim elements of patents to those relevant sections. This provides more accurate infringement analysis compared to using machine learning alone. The AI determines word meanings, searches element by element, and outputs matches with confidence scores.

Source

Method and apparatus for structuring patent claims to enable better search and classification of patent documents. The method involves parsing patent claims to extract syntactic components like preamble, relational phrases, patent components, connections, subject words, etc. It creates syntax rules by analyzing claims and defining phrases from ending tokens. Tags are attached to nouns and verbs before/after relational phrases to show relationships. The structured claims can be used for tasks like similar patent search and classification.

Source

Semantic patent document retrieval that improves accuracy and ease of use over existing methods. The method involves analyzing the semantic relationships between keywords in a patent document description. Instead of just searching for exact keyword matches, it determines a target keyword group and semantic relationships. Then, it generates search formulas based on these groups and relationships to find relevant documents. This allows more accurate and contextual retrieval compared to just exact keyword matching. The user can also refine searches by adjusting the target keywords and relationships.

Source

Automatic patent search to find similar patents more accurately and efficiently than manual methods. The search involves extracting keywords and classification numbers from a specific patent, prioritizing them based on manual input versus automated segmentation, and constructing a search formula using the prioritized terms until results meet thresholds. This improves accuracy by leveraging manual keywords and classifications as higher priority over automated ones. It also sorts search results by similarity to the original patent.

Source

Searching for information using concept expansion of a search term to improve efficiency compared to just using the initial search term. The concept expansion involves iteratively finding related terms and expanding the search query with them. This is done using techniques like sequential or parallel breadth-first search, or a neural network like Hopfield nets. The expanded query is then used to search the database for more relevant results.

Source

A patent retrieval method using keywords and multi-platform classification numbers to improve accuracy compared to traditional patent searching. The method involves obtaining competitive product information, retrieving patents for those products, extracting classification numbers and keywords from the patents, and using them to search for more relevant patents. This leverages existing patent data to determine better classification numbers and keywords for more accurate patent retrieval.

Source

Automated patent search system using machine learning to classify patents based on IPC standards and find similar patents. The system extracts keywords from patent documents, selects important keywords using a text rank algorithm, classifies IPC codes using a trained model, searches for comparable patents based on keywords and IPC, and analyzes topic similarity using LDA.

Source

Determining the validity or infringement of a patent using natural language processing and information retrieval techniques. The method involves indexing patent references to create search documents, building a search index from the documents, generating queries from patent claims, and executing searches in the index to find matching documents. This allows quickly identifying where in references a claim is disclosed. The searches use techniques like splitting references into smaller units, removing words, and substituting synonyms to improve accuracy.

Source

Judging patentability of a new invention based on comparison with existing patents. The method involves extracting a group of similar patents from a database, comparing the text of the new invention to the similar patents, and using metrics like edit distance and length index to determine patentability. The method aims to accurately evaluate whether an invention is novel and non-obvious by comparing it to existing patents.

Source

A research and development assistance system using patent data to suggest new ideas and overcome patent obstacles. The system analyzes patent documents using text mining to extract technical elements. Association rules are generated from these elements to identify combinations and strengths. Weak rule combinations are presented as R&D suggestions. Strong rule combinations are used to infer patent invalidations.

Source

A research and development (R&D) auxiliary system that leverages patent databases to provide suggestions and insights to R&D teams. The system uses association rules created by analyzing patent classification numbers or technical element messages from patent documents. These association rules, along with their strength values, are used to generate suggestions like patentable R&D ideas or patent invalidation inferences that can assist R&D efforts. By analyzing patents in this way, the system provides practical guidance based on the patent database that addresses the specific needs of R&D personnel.

Source

Semantic-based similar patent search technique that can calculate the similarity of technical terms by analyzing the usage of terms in patent literature. The method involves extracting key technical terms from patents, purifying the patents based on those terms, vectorizing the purified patents, and calculating cosine similarity between the vectors to find similarity between patents. This allows finding patents in different languages that belong to the same technical field.

Source

A method and apparatus for analyzing patent documents to extract meaningful keyword sequences that represent the technical content of the claims. The method involves breaking down the claims into constituent requirement units, like steps in a process, and identifying keywords within each unit. It then constructs a keyword sequence by connecting the keywords from adjacent requirement units. This sequence represents the technical flow and relationships between the steps in the claimed invention.

Source

Keyword suggestion method for efficient and accurate patent search. The method involves generating suggested keywords for patent search based on analysis of existing patent documents. It involves steps like statistical analysis of search results to identify important single words, integration of word counts, sorting, and keyword generation. The suggested keywords are then used in subsequent patent searches to improve efficiency and accuracy.

Source

Refining search results by leveraging concept markers to improve precision and flexibility in content searching. When a user submits a query, the system identifies relevant documents and concept markers associated with them. The user can then select concept markers to refine the search results. This allows refining by concept relevance instead of just keywords. The selected concept markers are re-ranked along with the documents to provide better results. The concept marker assignment involves identifying relevant concepts in documents and mapping them to master markers.

Source

A technology-convergent design solution business model that enables non-experts to easily search and utilize intellectual property information for design purposes. The business model involves a keyword search methodology that provides autocomplete suggestions, classified intellectual property databases, and keyword conversion to help users navigate the complex intellectual property search process. The method includes steps like retrieval expression autocomplete, search result providing, homogeneous/heterogeneous IP DB classification, and keyword conversion. By simplifying IP search and analysis for designers, it aims to provide a technology convergence design solution leveraging patented big data analysis.

Source

Integrated patent data mining system that enables efficient and accurate patent searching by extracting key phrases from user inputs using artificial intelligence algorithms, converting them into a single executable search query, and simultaneously searching multiple heterogeneous patent databases to provide search results. This improves patent data mining by leveraging AI to extract relevant search terms from user inputs and convert them into a single query for multi-database searching.

Source

A system for efficiently identifying potentially infringing products on a patent using a user interface and machine intelligence to extract keywords and formulate search queries based on context and intelligence from additional user input beyond just the patent text. This allows targeted searching across multiple databases through a single interface to identify infringing items.

Source

Automatic denoising method for patent retrieval to efficiently and accurately search large patent databases by combining keywords and classification numbers. The method involves determining a keyword from the target search document, obtaining a target database using that keyword, getting the first document from that database, determining the classification numbers within it, finding the sections, categories, etc, performing semantic analysis between those classifications and the target document, and deleting the document if the analysis doesn't match. This automatically removes irrelevant patents and improves search efficiency by focusing on relevant classifications.

Source

Automatically finding high-value patents by evaluating documents based on factors like claim count, word count, and patentee type. The method involves obtaining a target document containing a keyword, searching a patent database for matching documents, analyzing the documents' claim, word, and patentee data, and sending the best documents back to the database. This focused search and scoring approach improves accuracy and reduces false positives compared to broad keyword searches.

Source

Automatic patent retrieval method and device that improves accuracy and reduces noise in patent searches by expanding keywords and using multiple databases. The method involves obtaining a target search document with a keyword, expanding the keyword range, finding related documents in the expanded range, and then combining those databases to find the most relevant documents for the target search. This prevents missing relevant patents due to narrow initial keyword searches and reduces noise from unrelated documents.

Source

A method to retrieve patent information using association data to overcome the limitations of keyword searching. The method involves identifying and extracting associations between entities like inventors, organizations, and technologies from patent documents. These associations are then used as search queries to retrieve related patents beyond just keywords. This allows more comprehensive and effective retrieval of relevant patents for reference and analysis.

Source

Automating the harvesting and ranking of patent references to improve patent examination and analysis. Automated reference searching and analysis system that provides a tool to collect prior art references based on automatically harvested and pruned research threads. The system also ranks the harvested references for relevance to a claim set and presents the ranking visually.

Source

Patent search system that enables easily searching for patents by uploading technical documents without requiring complex search expressions. The system extracts data from the uploaded document, generates a retrieval expression from the extracted data, and searches patents based on that expression. The retrieval expression is created by selecting keywords from the extracted data based on factors like frequency and position. This allows users to find relevant patents by simply uploading their own documents instead of manually crafting complex search queries.

Source

A patent searching system that suggests keywords based on the structural attributes of patent claims. The system uses the user's search keyword to find patents, extracts definitions related to specific attributes (like subject matter, interrelationships), calculates occurrence frequencies, and suggests keywords from those definitions. This provides more precise search results by identifying keywords that are structurally related to the initial search term.

Source

A method to efficiently search for patent documents that are relevant to a given search criteria. The method involves incrementally increasing the set of conforming documents until a sensory evaluation passes. It uses formulas and keyword tables to find documents that match the search criteria. This is done by first creating formulas for finding conforming documents based on a master document set. Then, it calculates a comprehensive search formula for finding compatible documents. This involves creating formulas for each keyword layer and combining them. If the initial set of documents is too small, it randomly extracts more documents to meet a keyword sufficiency reference value. The goal is to find a set of documents that comprehensively covers the search criteria.

Source

Automatic method for extracting patent keywords from documents to improve the accuracy and comprehensiveness of patent searches. The method involves importing electronic documents, analyzing the title and text to retrieve similar documents, and exporting the final keyword list after contrasting.

Source

System for identifying prior art patent references using artificial intelligence that overcomes the limitations of traditional keyword search tools. The system preprocesses a corpus of patent references to identify keywords and n-grams (sequences of n words) along with statistics like occurrence counts. For a new patent application, it extracts n-grams from the claims section and ranks them based on the precomputed weights. It then finds prior art references matching the ranked n-grams and ranks them based on similarity. This provides more relevant prior art compared to keyword search.

Source

Intellectual property data acquisition and processing system that improves comprehensive patent searching and analysis by extracting synonyms related to keywords. The system has modules for keyword extraction, synonym analysis, and qualitative analysis. It connects to databases, thesauruses, and cloud resources to expand the search scope beyond exact matches. By extracting synonyms, it can capture variations in nomenclature across fields and improve patent data collection. This reduces the need for multiple searches and manual synonym matching.

Source