Information Search System Using Machine Reading Comprehension
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current information search engines often provide users with extensive and irrelevant results due to their inability to accurately understand user intent and deliver concise, relevant information, leading to user dissatisfaction in interactive searches.
Innovation Solution
An information search method and device that involves ranking webpages based on relevance, extracting and splicing text segments from relevant webpages, and using machine reading comprehension models to provide concise and accurate answers directly to the user, employing similarity calculation models and deep neural networks to improve search result relevance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If traditional search engines provide extensive search results based on keyword matching, then the quantity of information provided to users increases, but the relevance and conciseness of the information decreases
Solution Approach 1:
The patent extracts the most relevant information from search results by using natural language processing to identify and extract key entities, relationships, and answers directly from the content. Instead of presenting all search results, the system extracts only the essential information that directly answers the user's query, thereby improving relevance while reducing quantity.
Solution Approach 2:
The patent introduces an intermediary processing layer between the search engine and the user. This intermediary uses machine learning models and natural language understanding to analyze search results, extract meaningful information, and present it in a concise format. The intermediary acts as a mediator that transforms extensive raw search results into refined, relevant information.
2Loss of information
If search engines return multiple webpages and links for user exploration, then the comprehensiveness of information coverage increases, but the time required for users to find relevant information increases
Solution Approach 1:
The patent performs preliminary analysis and processing of search results before presenting them to users. The system pre-extracts key information, pre-ranks results based on relevance, and pre-processes content to identify the most important answers. This preliminary action saves users time by presenting ready-to-use information rather than requiring them to manually explore multiple pages.
Solution Approach 2:
The patent segments the extensive search results into distinct, organized sections such as direct answers, related entities, relationships, and additional context. By dividing the information into structured segments, the system allows users to quickly locate specific types of information without scanning through unorganized content, thereby reducing time loss while maintaining comprehensiveness.
3Productivity
If search engines use simple keyword matching and word frequency analysis, then the simplicity and speed of processing increases, but the understanding of true text meaning and user intent decreases
Solution Approach 1:
The patent changes the parameters of text analysis from simple keyword matching and word frequency counts to more sophisticated metrics including entity recognition, relationship extraction, semantic similarity, and contextual relevance. By changing these analysis parameters, the system achieves better understanding of text meaning while maintaining processing efficiency through optimized algorithms.
Solution Approach 2:
The patent replaces mechanical keyword matching systems with intelligent natural language processing systems that use machine learning models, semantic analysis, and contextual understanding. This substitution enables the system to comprehend true text meaning and user intent rather than merely counting word occurrences, thereby improving measurement precision while maintaining productivity through efficient AI processing.
Data Source
AI summary
An information search method is provided. The method includes: searching for webpages related to a search request through a search engine; extracting respective texts related to the search request from respective webpages and splicing the texts to obtain a spliced text; obtaining a text segment from the spliced text; and sending the obtained text segment to the search engine, to display the obtained text segment in an information search result through the search engine. The present disclosure can bring great advantages to the search engine in terms of user experience and interaction, and can satisfy user requirements for a function of an intelligent question and answer. Through present disclosure, it is beneficial to directly presenting a short text with higher relevance to the search request to the user, thereby saving time in screening information for the user.