Entity Record Creation via Attribute-Based Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for creating entity records are inefficient, requiring human intervention, leading to ambiguity in identifying similar entity names, inefficient data processing, and high time complexity, resulting in reduced performance and increased energy dissipation in data processing hardware.
Innovation Solution
A system and method that utilizes a computer system to receive unstructured data, structure it into entity records with associated attributes, bucket similar records, create intermediate clusters using a pattern recognition algorithm, calculate prediction probability scores, and combine records with scores above a threshold to create unambiguous entity records, thereby reducing human intervention and processing complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing techniques are used to create entity records by matching entity names, then entity records can be created, but ambiguous identification of similar entity names occurs and human intervention is required
Solution Approach 1:
The patent transforms entity identification from simple name-matching to multi-parameter analysis by extracting and comparing entity attributes (type, category, properties) in addition to names. This parameter expansion enables automated distinction between similar entities without human intervention.
Solution Approach 2:
The patent replaces manual human verification with an automated computational system that uses pattern recognition algorithms and attribute-based comparison to resolve entity ambiguities, eliminating the need for human operators while improving consistency.
2Speed
If existing techniques search for entity records by matching entity names, then entity records can be retrieved, but the speed of response is insufficient
Solution Approach 1:
The patent performs preliminary structuring of unstructured data into standardized entity records with extracted attributes before the actual search operation. This pre-processing organizes data in a query-friendly format, enabling faster response times during retrieval operations.
Solution Approach 2:
The patent segments entity data into distinct components (entity name, type, attributes, properties) that can be independently processed and queried. This segmentation allows the system to search specific attribute fields rather than scanning entire records, improving search speed.
3Productivity
If existing techniques create entity records with ambiguous data, then entity records are created, but considerable data processing and time complexity are involved
Solution Approach 1:
The patent implements a streamlined data flow system where unstructured data is automatically structured, entities are extracted and attributed in a single pass, and records are immediately ready for use. This continuous automated pipeline eliminates manual verification steps and reduces processing time complexity.
4Reliability
If existing techniques require human intervention to verify and rectify errors, then entity records can be corrected, but human effort and time are consumed
Solution Approach 1:
The patent implements self-verification mechanisms where the system automatically validates extracted entity attributes against predefined schemas and cross-checks consistency within the entity record structure. This self-service validation catches errors automatically without requiring human operators.
Data Source
AI summary
A system that creates entity records, wherein the system includes a computer system. The system includes: a database arrangement operable to store existing data sources and a processing module communicably coupled to the database arrangement. The processing module is operable to receive unstructured data from existing data sources; structure the unstructured data to obtain data-records, wherein the data-records comprise entity names and entity attributes associated with the entity names; bucket data-records into one or more buckets based on similar entity attributes thereof; create intermediate clusters including at least one of the one or more buckets based on a pattern recognition algorithm; calculate a prediction probability score associated with each of the data-records within the intermediate clusters; combine the data-records from the intermediate clusters to obtain an entity-record, wherein the combined data-records include a prediction probability score higher than a predefined threshold.