User-Specific Semantic Relatedness Dataset Generation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for determining semantic relatedness in natural language processing rely heavily on unsupervised learning and knowledge engineering, which may not accurately capture user-specific or context-dependent semantic relationships.

Innovation Solution

A computerized method and system that calculates the semantic relatedness of terms by analyzing the co-appearance of pairs of terms in text segments associated with a targeted user, using weighted prevalence to evaluate and generate a user-specific semantic relatedness dataset, which can be used for personalized content aggregation, search, and advertising.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If unsupervised learning and knowledge engineering procedures are used to extract semantic information from background knowledge corpus, then the system can operate without user-specific data, but the accuracy of capturing user-specific and context-dependent semantic relationships deteriorates

Engineering Contradiction:

Improveuser-specific semantic relationship captureVSAvoidsemantic relatedness accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system performs preliminary actions by collecting and storing user-specific text segments (webpages visited, documents retrieved, emails sent/received, files stored) before semantic analysis is needed. This preliminary data collection enables the system to later generate accurate user-specific semantic relatedness without requiring real-time user input during the analysis phase.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system applies local quality by creating user-specific semantic relatedness datasets tailored to each individual user's characteristics, rather than using a single universal semantic model. Each user receives customized semantic analysis based on their personal text segments, ensuring high accuracy for user-specific contexts while maintaining the ability to generalize across different users.

Inventive Principle:

Principle #3Local quality

2Measurement precision

If supervised learning with user-specific text segments is implemented, then semantic relatedness accuracy improves, but system complexity and data processing requirements worsen

Engineering Contradiction:

Improvesemantic relatedness accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the overall semantic analysis task into distinct components: text segment collection, text segment weighting, co-appearance calculation, and semantic relatedness evaluation. This segmentation allows each component to be optimized independently and processed in manageable stages, reducing overall system complexity while maintaining high accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system introduces an intermediary semantic relatedness dataset that serves as a bridge between raw user-specific text segments and final semantic analysis results. This intermediary dataset pre-processes and structures the information, making subsequent analysis simpler and more efficient while preserving user-specific characteristics.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Productivity

If traditional unsupervised methods are used, then processing speed is maintained, but the ability to capture context-dependent semantic relationships deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoidcontext-dependent semantic relationship capture

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system changes key parameters by introducing user-specific weighting factors and context-dependent coefficients into the semantic relatedness calculation. These parameter modifications enable the system to capture context-dependent relationships while maintaining efficient processing through optimized mathematical formulations that leverage the pre-structured user-specific datasets.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS8909648B2Methods and systems of supervised learning of semantic relatedness

Publication Date: 2014.12.09 TECHNION RES & DEV FOUND LTD

AI summary

A method of evaluating a semantic relatedness of terms. The method comprises providing a plurality of text segments, calculating, using a processor, a plurality of weights each for another of the plurality of text segments, calculating a prevalence of a co-appearance of each of a plurality of pairs of terms in the plurality of text segments, and evaluating a semantic relatedness between members of each the pair according to a combination of a respective the prevalence and a weight of each of the plurality of text segments wherein a co-appearance of the pair occurs.