Multi-Granularity Machine Learning for Textual Content Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current approaches for generating suggestions in opportunity platforms rely heavily on string matching without considering the context of textual content, leading to irrelevant suggestions that worsen the opportunity creation process for providers.
Innovation Solution
A system and method using machine learning techniques to analyze items in textual content at multiple levels of granularity (sentence, segment, and document levels) to determine relevant suggestions by leveraging learned entity embeddings and machine-learned models, improving the quality of recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If string matching is used to generate suggestions, then the process is simple and fast, but the suggestions are irrelevant and do not consider context
Solution Approach 1:
The system segments the document into multiple levels of granularity including sentences, segments, and the full document. Each level is analyzed separately by machine learning models to extract contextual information, allowing the system to generate more relevant suggestions while managing complexity through hierarchical processing
Solution Approach 2:
The system transitions from single-dimensional string matching to multi-dimensional analysis by considering multiple levels of granularity (sentence level, segment level, document level) and combining results from different machine learning models, thereby improving suggestion relevance without overwhelming complexity
2Measurement precision
If machine learning analysis at multiple levels is used, then suggestion relevance is improved, but processing time and computational resources increase
Solution Approach 1:
The system performs preliminary analysis at the sentence level first, identifying potential items and their contexts before proceeding to segment and document level analysis. This staged approach allows early filtering and reduces the computational burden of more intensive analysis on the full document
Solution Approach 2:
The system applies machine learning analysis at multiple levels of granularity, performing more analysis than a simple string matching approach would require. By analyzing sentences, segments, and the full document, the system ensures high accuracy in item identification while the hierarchical structure prevents complete processing of all possible combinations
Data Source
AI summary
Techniques are provided for using machine learning techniques to analyze textual content. In one technique, a potential item is identified within a document. An analysis of the potential item is performed at multiple levels of granularity that includes two or more of a sentence level, a segment level, or a document level. The analysis produces multiple outputs, one for each level of granularity in the multiple levels of granularity. The outputs are input into a machine-learned model to generate a score for the potential item. Based on the score, the potential item is presented on a computing device. In response to user selection of the potential item, an association between the potential item and the document is created. The association may be used later to identify a set of users to which the document (or data thereof) is to be presented.