Concept-Based Search Classification for Abstract Natural Language Discourse
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current keyword-based search methods are inadequate for retrieving records that express or utilize high-level, abstract concepts, as they rely on specific keywords that are difficult to define, leading to large and irrelevant search results, and popularity metrics do not accurately represent the relevance of search topics like new uses of technologies or technologies addressing specific problems.
Innovation Solution
The implementation of concept-based search and classification techniques that analyze natural language discourse to identify and rank records based on pinnacle concepts, using a two-phase approach combining keyword-based and concept-based searching, where concept-based databases are created during the indexing phase and used in conjunction with keyword-based searching to improve relevance and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If keyword-based search is used, then search coverage is improved, but search precision deteriorates due to large and irrelevant result sets
Solution Approach 1:
The patent segments the search process into two distinct phases: keyword-based search to retrieve candidate records and concept-based classification to filter and rank them. This segmentation allows the system to first cast a wide net with keywords (maintaining coverage) then apply precise conceptual filtering (improving precision), resolving the contradiction between comprehensive search coverage and accurate result retrieval.
Solution Approach 2:
The patent introduces concept-based classification as an intermediary mechanism between keyword search and final result presentation. This intermediary layer analyzes linguistic features and contextual relationships to bridge the gap between simple keyword matching and meaningful concept understanding, thereby improving search precision without sacrificing the broad coverage achieved through keyword-based retrieval.
2Measurement precision
If specific keywords are used, then search precision is improved, but search coverage deteriorates due to inability to capture abstract concepts
Solution Approach 1:
The patent adds a new dimension to search by transitioning from one-dimensional keyword matching to multi-dimensional concept analysis. By incorporating linguistic feature analysis, contextual relationships, and semantic understanding, the system captures abstract concepts that cannot be expressed through specific keywords alone, thereby expanding search coverage while maintaining precision through concept-based filtering.
Solution Approach 2:
The patent changes the search parameters from simple keyword presence/absence to complex conceptual attributes including linguistic features, contextual relationships, and semantic meanings. This parameter transformation enables the system to identify records expressing abstract concepts like 'new uses of technology' or 'technologies addressing problems' without relying on specific keywords, thus improving coverage while maintaining precision.
3Ease of operation
If popularity metric is used for ranking, then ease of operation is improved, but measurement precision deteriorates for concept-based search topics
Solution Approach 1:
The patent changes the ranking parameter from popularity-based metrics to concept-based relevance scoring. By analyzing whether records actually express or utilize the target concepts through linguistic feature examination and contextual analysis, the system achieves precise relevance ranking for concept-based search topics, replacing the simplistic popularity metric with a nuanced conceptual relevance measure.
Data Source
AI summary
Pinnacle concepts are not amenable to detection by the use of keywords. A unit of natural language discourse (UNLD) “refers” to a pinnacle concept “C” when that UNLD uses linguistic expressions in such a way that “C” is regarded as expressed, used or invoked by an ordinary reader of “L.” A reference can have a “reference level” value that is proportional to: the “strength” with which the pinnacle concept is referenced, the probability that a pinnacle concept is referenced or both strength and probability. Pinnacle concepts can be divided into Quantifiers and non-Quantifiers. A Quantifier can modify the reference level assigned to a non-Quantifier. A concept “C,” that is determined to be referenced by a UNLD “x,” after application of its Quantifiers, is said to be asserted by “x.” Concept-based classification is the identification of whether a pinnacle concept “C” is asserted by a UNLD. Concept-based classification can be used for concept-based search.