Document Chunk Identification for Search Context
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search engines struggle to efficiently locate and highlight relevant information within documents, requiring users to open documents and perform additional searches, which is time-consuming and inefficient, especially when search keywords are not in the exact order or are scattered across different chunks.
Innovation Solution
A computer system that processes user search keywords by identifying and highlighting relevant chunks within primary and secondary documents, using a query engine, cache engine, and stream engine to retrieve and semantically annotate document content, allowing for efficient chunk identification and display without the need to open documents.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional search engines return document names with one or two matching sentences, then search results are provided quickly, but users cannot understand the meaning of search keywords in context and must open documents to find relevant information
Solution Approach 1:
The patent divides documents into semantic chunks (paragraphs, sections, or meaningful units) and identifies relevant chunks that contain search keywords. Instead of returning entire documents or random snippets, the system segments and returns only the specific chunks that are relevant to the search query, making it easier for users to find information without opening full documents.
Solution Approach 2:
The patent introduces an intermediary chunk identification system between the search engine and the user. This intermediary analyzes document structures, identifies semantic chunks, and presents relevant portions to users before they need to open full documents. The chunk identification acts as a mediator that bridges the gap between quick search results and contextual understanding.
2Loss of information
If users open documents using native applications to understand search results, then they can access full document content, but this is time-consuming and requires additional search operations within the document
Solution Approach 1:
The patent performs preliminary chunk identification and relevance analysis before the user needs to access the document. By pre-identifying and marking relevant chunks within documents during the search process, the system prepares the information in advance, so users receive ready-to-view relevant portions without needing to open documents and perform additional searches.
Solution Approach 2:
The patent extracts and presents only the relevant chunks containing search keywords from within full documents. Instead of requiring users to access entire documents, the system takes out and displays only the specific portions that are relevant to the search query, reducing the time and effort needed to find information while maintaining access to complete relevant content.
3Measurement precision
If search tools require keywords to be in exact order to match, then search precision is maintained, but users may miss relevant information where keywords appear in different orders
Solution Approach 1:
The patent implements dynamic chunk identification that adapts to different keyword arrangements within semantic units. Rather than requiring fixed keyword sequences, the system dynamically identifies chunks that contain relevant keywords regardless of their order, allowing flexible matching while maintaining precision through semantic context analysis.
Solution Approach 2:
The patent changes the matching parameter from strict sequential keyword order to semantic chunk relevance. By shifting the matching criterion from positional accuracy to semantic relevance within identified chunks, the system maintains search precision while becoming adaptable to different keyword arrangements and document structures.
4Ease of operation
If users limit searches to a subset of keywords to reduce complexity, then search operation becomes simpler, but many instances of the subset may appear in documents requiring significant effort to find relevant content
Solution Approach 1:
The patent segments documents into meaningful chunks and applies chunk-level identification even when users search with limited keywords. By dividing documents into searchable semantic units and identifying relevant chunks among multiple matches, the system reduces the effort needed to locate relevant information while maintaining simple search operations.
Data Source
AI summary
A computer receives a request to search one or more secondary documents. At least one of the secondary documents is associated with a primary document. The computer searches at least a subset of the secondary documents for documents that satisfy the search request and identifies at least one secondary document that satisfies the search request.