Keyword Association Generation Using Deep Learning NER and Co-occurrence Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating keyword information related to financial data lack efficiency in identifying highly associated keywords, especially in unstructured finance-related text data, leading to incomplete or irrelevant results.
Innovation Solution
An electronic device employs a deep learning-based named entity recognition (NER) model to identify keywords and determine their degrees of association, generating response words by analyzing co-occurrence graphs and filtering based on association degrees, using morpheme analysis for additional keyword classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional keyword extraction methods are used on unstructured finance-related text data, then the processing is simpler and faster, but the identified keywords lack high association with query words and investment themes
Solution Approach 1:
The patent segments the keyword extraction process into multiple specialized modules: NER model for entity recognition, co-occurrence graph construction for relationship mapping, and degree of association calculation for ranking. Each module handles a specific aspect of keyword analysis, improving overall precision while managing complexity through functional decomposition
Solution Approach 2:
The patent introduces a co-occurrence graph as an intermediary structure between the raw text data and the final keyword results. This graph serves as a mediator that captures relationships between keywords and query words, enabling more accurate association measurement without directly complicating the extraction process
2Measurement precision
If deep learning-based NER model and co-occurrence graph analysis are employed to identify highly associated keywords, then the keyword relevance improves, but the processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary action by pre-processing the unstructured text data to extract entities and construct the co-occurrence graph before the actual keyword query processing. This pre-computation of relationships allows faster retrieval and analysis when specific keywords are queried, reducing real-time processing time while maintaining high relevance accuracy
Solution Approach 2:
The patent applies partial action by focusing the deep learning NER model and co-occurrence analysis only on the specific query word and its related entities, rather than analyzing the entire text corpus for every query. This selective approach maintains high keyword relevance while significantly reducing processing time and computational overhead
Data Source
AI summary
Provided is a method of generating keyword information by an electronic device. The method includes identifying a text set including at least one text element. The method includes using a named entity recognition (NER) model based on deep learning, identifying keywords in the at least one text element. The method includes based on the text set, determining degrees of association of keyword pairs included in a keyword set including the keywords. The method includes obtaining information on a query word that is input by a user. The method includes based on at least one among the degrees of association, generating information on response words corresponding to the query word.