Synonym Retrieval via Interactive Feedback and TF-IDF Scoring

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing data exploration methods fail to effectively identify synonyms in text documents, particularly missing abbreviations, misspellings, and variations, which limits the retrieval of relevant information.

Innovation Solution

A method and apparatus that input a key word, present possible synonyms, solicit user feedback, determine match scores, and retain a predetermined number of top synonyms, discarding lower-scored synonyms, using TF-IDF vectors and inverted indexes for efficient synonym finding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional keyword search is used to find synonyms, then the search process is simple and fast, but it fails to identify abbreviations, misspellings, and variations of the key word

Engineering Contradiction:
Improvesynonym identification accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary human user who provides feedback on synonym candidates. The system presents potential synonyms to the user, who then validates or rejects them. This human intermediary bridges the gap between automated text processing and accurate synonym identification, allowing the system to learn and improve without requiring complex pre-programmed linguistic rules

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements a feedback loop where user responses to synonym candidates are collected and used to update the inverted index and match scoring. Each user interaction provides feedback that refines the system's understanding of synonym relationships, progressively improving accuracy while maintaining a relatively simple system architecture

Inventive Principle:
Principle #23Feedback

2Measurement precision

If all possible synonyms are retained for a key word, then comprehensive coverage is achieved, but data exploration efficiency decreases due to information overload

Engineering Contradiction:
Improvesynonym retrieval completenessVSAvoiddata exploration efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system changes the parameter of synonym selection by introducing match scores based on TF-IDF calculations and user feedback. Instead of retaining all possible synonyms, the system ranks them by relevance score and presents only the top candidates to users. This parameter-based filtering maintains comprehensive coverage potential while improving efficiency by focusing user attention on the most relevant synonyms

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The system performs partial action by initially presenting only a subset of potential synonyms (those with highest match scores) rather than all possible synonyms. This partial presentation is sufficient for most cases, and the system can expand to show more candidates if needed, balancing completeness with efficiency

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS8812297B2Method and system for interactively finding synonyms using positive and negative feedback
Publication Date: 2014.08.19 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US8812297B2 patent drawing
  • US8812297B2 patent drawing
  • US8812297B2 patent drawing

AI summary

Determining synonyms of words in a set of documents. Particularly, when provided with a word or phrase as input, in exemplary embodiments there is afforded the return of a predetermined number of “top” synonym words (or phrases) for an input word (or phrase) in a specific collection of text documents. Further, a user is able to provide ongoing and iterative positive or negative feedback on the returned synonym words, by manually accepting or rejecting such words as the process is underway.