Phrase Concept Search Using Phonetic Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical data retrieval systems using natural language processing face challenges in analyzing and normalizing varying sentence forms that express the same concept, leading to incorrect search results due to differences in word usage and sentence structure.
Innovation Solution
The system analyzes documents into sentences, keeps word position information, and generates comprehensive logical queries by deriving synonyms and semantic equivalents, ensuring relevant documents are retrieved by considering words' presence within the same sentence and applying phonetic representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If natural language processing is used to allow doctors to write prognosis in their own words, then ease of operation is improved, but manufacturing precision deteriorates due to variations in sentence forms expressing the same concept
Solution Approach 1:
The system changes the parameter of word representation by using phonetic transcriptions (e.g., Metaphone algorithm) to convert words into sound-based representations. This allows semantically equivalent words with different spellings or forms to be recognized as matching, thus maintaining precision while allowing natural language variation.
Solution Approach 2:
The patent introduces an intermediary normalization layer between the natural language input and the search matching process. This intermediary layer standardizes sentence structures and word forms while preserving the original meaning, enabling both ease of writing and precise matching.
2Device complexity
If search engines define maximum distance between words in query, then device complexity is reduced, but measurement precision deteriorates because words may not be in the same sentence
Solution Approach 1:
The patent segments the search process into distinct phases: first breaking down the query into individual words, then searching for each word while enforcing the constraint that all words must appear in the same sentence. This segmentation allows the system to maintain algorithmic simplicity while improving precision through the sentence-level constraint.
3Productivity
If traditional search algorithms are used without sentence structure analysis, then productivity is improved through faster processing, but reliability deteriorates due to wrong results being retrieved
Solution Approach 1:
The patent performs preliminary action by pre-processing and normalizing the query phrase before the actual search execution. This includes analyzing the sentence structure, identifying key concepts, and preparing normalized forms of the query terms in advance, which enables both reliable results and efficient processing during the actual search.
Data Source
AI summary
A system and method for fast concept search in multiple documents where the concept is expressed by plurality of words, all of which have to be in the same sentence and within specified range. The system automatically finds equivalent expressions of the same concept, and returns as search results all documents in which the concept is contained.