An Analysis of Proposition Abstraction from Domain Text for Conceptual Plan Mining Using Deep Learning Methods

Summarization is a perfect example of one of the numerous text mining jobs that may be improved with concept maps. A broad idea map illuminates the connections between the numerous thoughts being mapped. This article describes an evaluation technique designed with tokenization, proposition extraction, and refining in mind. The Deep Tokenizer approach, which is based on deep learning principles, has a processing time of 5 milliseconds and a tokenization accuracy of 98%. The extraction accuracy with the given proposition extractor approach was 95%, and the processing time was 10 milliseconds. The program was modified specifically to extract propositions from tokenized text. The suggested proposition refiner technique was proven to have a refining accuracy of 94% and a processing time of 8 milliseconds. The goal of this strategy was to back up and enhance the ideas derived from the data. Bar charts were used to graphically show the performance statistics of the various algorithms. These findings indicate the techniques' usefulness and reliability in dealing with large text datasets, shedding light on possible applications in natural language processing.