Patent Search Engine Using Citation Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing volume of patent-related literature makes it difficult for professionals to efficiently search for relevant information, especially when evaluating the novelty, validity, and value of patent assets, due to the reliance on rudimentary keyword matching functions and the need for experienced-based search strategies.
Innovation Solution
A Patent Related Publication Search Engine (PRPSE) is developed to perform iterative searches using citation analysis, determining predominant keywords, and suggesting related keywords and categories, enabling users to refine search results through a user-friendly interface and API, facilitating the identification of relevant patent-related publications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the corpus of patent and scientific literature is expanded to include more sources, then the completeness and accuracy of search results is improved, but the difficulty and time required to search through the information increases
Solution Approach 1:
The system segments the large corpus of patent and scientific literature into manageable units by processing documents individually through automated extraction and analysis pipelines. Each document is broken down into structured data elements (titles, abstracts, claims, keywords) that can be independently indexed and searched, making the vast corpus accessible without requiring manual review of entire documents.
Solution Approach 2:
The system introduces an intermediary automated search engine that acts as a mediator between the user and the vast corpus of literature. This intermediary performs automated keyword extraction, citation analysis, and relevance ranking, translating user queries into precise search results without requiring users to manually navigate through the entire corpus.
2Productivity
If automated search tools are used to process large volumes of patent literature, then productivity is improved, but the precision and nuance of search results may deteriorate compared to expert human searchers
Solution Approach 1:
The system implements feedback mechanisms where search results are continuously refined based on citation patterns and relevance metrics. The automated extraction engine learns from citation relationships and user interactions, adjusting its keyword extraction and ranking algorithms to improve precision while maintaining high productivity. This feedback loop allows the system to converge toward expert-level accuracy.
Solution Approach 2:
The system combines multiple automated analysis techniques (keyword extraction, citation analysis, semantic matching) into a composite search engine that leverages the strengths of each method. By integrating these different approaches, the system achieves both high productivity through automation and high precision through the complementary nature of the combined techniques.
3Measurement precision
If professionals rely on experience-based search strategies and trial-and-error methods, then the quality of search formulation is improved, but the time and expertise required to perform searches increases
Solution Approach 1:
The system enables self-service searching by automatically analyzing the input query and generating optimized search strategies without requiring user expertise. The automated extraction engine identifies relevant keywords, suggests search terms, and formulates search queries based on the content being searched, allowing any user to perform expert-level searches without needing to learn complex search strategies.
Solution Approach 2:
The system dynamically changes search parameters (keywords, weightings, filters) based on the specific query and corpus being searched. Rather than requiring users to manually adjust these parameters based on experience, the system automatically optimizes them by analyzing the input text and adapting the search strategy in real-time, simplifying the process while maintaining high quality results.
Data Source
AI summary
Methods, systems, and techniques for facilitating searching for patent related literature are provided. Some examples provide a Patent Related Publication Search Engine (“PRPSE”), which enables users and or programs through the use of an application programming interface (“API”) to iteratively find patent related publications such as issued patents and patent application publications. In typical operation, based upon input text, the PRPSE determines predominant keywords found in the text and locates a set of patent related publications most closely related to these determined keywords as search results. In some examples, the PRPSE determines the set of patent related publications most closely related to these determined keywords both by full text searching of a corpus of patent related publications for all of the patent related publications that contain the predominant keywords and by finding all correlated patent related publications through citation analysis.