Sentence Search Apparatus Using Dependency Relations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sentence search systems often struggle to effectively narrow down search results using recommended words, as these words are determined solely by the degree of appearance of each word in the search results, which may not accurately reflect the user's intent.
Innovation Solution
A sentence search apparatus that includes an acceptance unit, a search unit, an analysis unit, and a determination unit, which performs sentence analysis to determine dependency relations between words and selects recommended words based on these relations, along with the inquiry sentence and search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If recommended words are determined based on the degree of appearance of each word in search results, then the system is simple to implement, but the recommended words do not accurately reflect user intent and fail to effectively narrow down search results
Solution Approach 1:
The patent changes the parameter for determining recommended words from simple word appearance frequency to dependency relation strength. By analyzing grammatical dependencies between words in the inquiry sentence and search result sentences, the system identifies words that have strong relational ties to the user's intent, thereby improving the accuracy of recommended words while maintaining system feasibility
Solution Approach 2:
The patent introduces sentence analysis as an intermediary process between the search engine and the recommended word selection. This intermediary analyzes the grammatical structure and dependency relations within sentences, enabling the system to select recommended words based on their structural relationship to the inquiry rather than just their frequency of appearance
2Measurement precision
If sentence analysis is performed to determine dependency relations between words, then the accuracy of recommended words improves, but the computational complexity and processing time increase
Solution Approach 1:
The patent segments the sentence analysis process into focused dependency relation extraction rather than comprehensive grammatical analysis. By concentrating only on identifying dependency relations between specific words in the inquiry and search results, the system achieves improved accuracy without the full computational overhead of complete sentence parsing
Solution Approach 2:
The patent applies partial action by performing dependency analysis only on relevant portions of sentences - specifically focusing on words that appear in both the inquiry and search results. This selective approach extracts sufficient dependency information to improve recommended word accuracy without analyzing the entire sentence structure, thereby controlling computational complexity
Data Source
AI summary
A sentence search apparatus includes a processing circuit. The processing circuit: accept a character string input by a user; search a sentence collection in which a plurality of sentences as a search target is accumulated based on the accepted character string to acquire one or more search result sentences matching the accepted character string; perform sentence analysis on the search result sentence to acquire a dependency relation between words constituting the search result sentence; and determine one or more recommended words for narrowing down the search result sentence to a sentence desired by the user based on the accepted character string, the search result sentence, and the dependency relation.