Expression Clustering for Same Notation Distinction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information processing systems struggle to distinguish between expressions with the same notation but different meanings, leading to difficulties in handling related expressions that may or may not be suitable for user preferences during content recommendation.
Innovation Solution
An information processing apparatus and method that extracts and clusters expressions based on feature amounts, calculates assignment degree vectors, and identifies related expressions with similar vectors to provide accurate and differentiated search results, including a synonym extraction unit to integrate scores and cluster components for improved recommendation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If expressions having the same notation are handled as one expression, then the system complexity is reduced, but the ability to distinguish different meanings is lost
Solution Approach 1:
The patent segments expressions having the same notation into multiple clusters based on their different meanings. Each cluster represents a distinct semantic category, allowing the system to maintain separate handling for expressions with identical notations but different contexts or meanings. This segmentation resolves the contradiction by dividing the unified expression handling into meaningful subgroups.
Solution Approach 2:
The patent applies local quality by assigning different properties (cluster assignments) to different parts (expressions) of the same notation based on their contextual meanings. Each expression is characterized by its cluster assignment vector, which captures its specific semantic context. This allows the system to treat expressions with the same notation differently according to their local semantic quality.
2Speed
If related expressions are searched based on similarity in feature amount vectors, then the search speed is improved, but the accuracy of distinguishing different meanings is reduced
Solution Approach 1:
The patent introduces a new dimension (cluster assignment vectors) to the existing feature amount vectors. Instead of relying solely on traditional feature similarity, the system incorporates cluster assignments as an additional dimension for comparison. This allows the search to consider both the original feature similarity and the semantic clustering information, improving meaning distinction accuracy while maintaining search efficiency through vector-based operations.
3Measurement precision
If expressions are clustered into multiple groups, then the distinction of different meanings is improved, but the processing complexity increases
Solution Approach 1:
The patent performs preliminary clustering of expressions into multiple groups before the actual search operation. By pre-computing the cluster assignments and storing them as assignment vectors, the system prepares the data structure in advance. This preliminary action reduces the processing complexity during search operations, as the clustering work is done once beforehand rather than repeatedly during each search.
Solution Approach 2:
The patent creates a copied representation (cluster assignment vector) of each expression that captures its semantic grouping information. Instead of performing complex clustering operations during each search, the system uses these pre-computed copied representations to quickly compare and distinguish expressions with the same notation but different meanings, reducing processing complexity while maintaining distinction accuracy.
Data Source
AI summary
Disclosed is an information processing apparatus including an expression extraction unit, a feature extraction unit, a clustering unit, a related expression extraction unit, and an output unit. The expression extraction unit extracts a plurality of expressions from a plurality of documents. The feature extraction unit extracts feature amounts of the extracted respective expressions while distinguishing the expressions having the same notation. The clustering unit clusters the extracted respective expressions together while distinguishing the expressions having the same notation and calculates assignment degree vectors having assignment degrees of the respective expressions to two or more respective clusters as components. The related expression extraction unit extracts related expressions having the assignment degree vectors similar to those of a provided input expression while distinguishing the expressions having the same notation. The output unit outputs the related expressions and identification information for identifying the related expressions.