Keyword Input Support Using Semantic Attribute Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information retrieval methods require a pointing device for keyword selection, limiting their application to devices without pointers, such as cellular phones and television sets, and only allow retrieval of information related to the currently viewed document, not past documents.
Innovation Solution
A keyword input supporting apparatus and method that acquires and analyzes text data from documents, selects main components, applies part-of-speech and semantic attribute analysis, extracts specific names, and classifies them as keyword candidates for presentation to the user, enabling keyword input without a pointing device.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a pointing device is used for keyword selection, then retrieval accuracy is improved, but device compatibility deteriorates
Solution Approach 1:
The patent replaces the mechanical pointing device interaction with automatic text analysis and keyword extraction algorithms. The system automatically identifies and extracts keywords from document text using part-of-speech analysis and semantic attribute detection, eliminating the need for manual pointing device selection while maintaining retrieval accuracy.
2Measurement precision
If only current document keywords are used, then retrieval relevance is improved, but information completeness deteriorates
Solution Approach 1:
The patent performs preliminary keyword extraction and storage from multiple documents before retrieval operations. The system pre-processes documents to extract keywords and stores them in a database, enabling comprehensive search across multiple documents without requiring re-analysis during retrieval, thus maintaining both relevance and completeness.
Solution Approach 2:
The patent creates a universal keyword extraction system that can process and extract keywords from multiple different documents simultaneously. The extracted keywords from various documents are stored together in a unified database, allowing the system to perform comprehensive retrieval across multiple documents using the same mechanism.
3Measurement precision
If manual keyword selection is required, then retrieval precision is improved, but operation complexity deteriorates
Solution Approach 1:
The patent implements a self-service keyword extraction system where the computer automatically performs keyword identification and extraction without requiring user intervention. The system uses natural language processing to automatically analyze document text, identify keywords based on semantic attributes and part-of-speech analysis, and prepare them for retrieval operations, thereby simplifying user operations while maintaining retrieval precision.
Data Source
AI summary
A keyword input supporting apparatus includes a document acquisition unit that acquires a document having a plurality of components containing text data, a main component selection unit that selects a component having many characters in the text data as a main component, a part-of-speech analysis unit that analyzes the part-of-speech of the text data contained in the main component, and adds a semantic attribute to each of words of the text data, a specific name extraction unit that extracts as a specific name a word, having a predetermined semantic attribute or part of speech, from the words, a specific name storage that stores the specific name together with the corresponding semantic attribute, a keyword candidate classification unit that performs classification of the specific name from the storage as a keyword candidate based on the semantic attribute, and a keyword candidate presentation unit that presents the keyword candidate to a user.