Document Clustering for Information Exploration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information retrieval methods, such as term searching with Boolean operations, are sensitive to search term choices and fail to account for synonyms, misspellings, and differences in term usage, leading to irrelevant results and requiring user familiarity with the subject matter.
Innovation Solution
The method employs clustering techniques, specifically latent semantic analysis and singular value decomposition, to group documents by concept space vectors, allowing users to explore document sets hierarchically and identify themes, thereby reducing the need for user familiarity with the subject matter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If term searching with Boolean operations is used, then computational efficiency is improved, but search result relevance deteriorates due to sensitivity to search term choices and failure to account for synonyms and different word forms
Solution Approach 1:
The patent introduces an intermediary layer (the thesaurus and concept hierarchy) between the user's search terms and the document database. Instead of directly matching search terms to documents, the system translates search terms into conceptual representations using the thesaurus, then performs searching based on these conceptual mappings. This intermediary mechanism resolves the contradiction by maintaining computational efficiency while significantly improving search result relevance through synonym and related term expansion.
Solution Approach 2:
The patent transforms the search parameter from literal term matching to conceptual similarity matching. By changing the parameter from exact word matching to concept-based matching using the thesaurus and concept hierarchy, the system overcomes the limitation of being sensitive to specific word choices while maintaining computational feasibility through structured conceptual relationships.
2Extent of automation
If term searching with Boolean operations is used, then automated searching capability is improved, but user familiarity with subject matter and searching tools is required, increasing operation difficulty
Solution Approach 1:
The patent implements self-service by enabling the system to automatically perform concept-based expansion and translation without requiring user expertise. The thesaurus and concept hierarchy automatically expand search terms into related concepts and synonyms, allowing users to perform effective searches using simple terms without needing to understand the subject matter deeply or master complex search syntax.
Solution Approach 2:
The thesaurus acts as an intermediary that bridges the gap between user-friendly simple search terms and the complex structured document database. This intermediary automatically handles the translation and expansion, making the searching process easier for users while maintaining high automation capability.
3Reliability
If comprehensive term matching is performed to account for synonyms and different word forms, then search result relevance is improved, but computational complexity increases
Solution Approach 1:
The patent segments the comprehensive term matching task into manageable components by organizing terms into a hierarchical concept structure with thesaurus relationships. Instead of performing exhaustive comparisons across all possible term variations, the system divides the search space into conceptual categories and relationships, reducing computational complexity while maintaining comprehensive coverage of synonyms and related terms.
Solution Approach 2:
The patent performs preliminary action by pre-organizing the thesaurus and concept hierarchy before searching. The conceptual relationships, synonym mappings, and hierarchical structures are established in advance, allowing the search process to efficiently traverse pre-computed relationships rather than performing complex real-time analysis of all term variations.
Data Source
AI summary
Disclosed information exploration system and method embodiments operate on a document set to determine a document cluster hierarchy. An exclusionary phrase index is determined for each cluster, and representative phrases are selected from the indexes. The selection process may enforce pathwise uniqueness and balanced sub-cluster representation. The representative phrases may be used as cluster labels in an interactive information exploration interface.