Technique of Semantic Unambiguity for a Concept Selection of Terms in Focused Contexts with Reinforcement Learning Integration

Abstract

Nowadays, there have been many developments of learning processes for computers to understand the meaning of words and their semantic similarities in order for the computers to better communicate, interact and exchange information with humans. Semantic learning development is a major issue because computers cannot comprehend the suitable meaning of words in the concerning concept. As a result, this research is proposing and exploring the efficiency of the technique of semantic unambiguity in order to clarify the Term Concepts in the focused contexts. From the case study with 22 contexts, 62 term, and 475 synsets, it was shown that Reinforcement Learning could accurately select the suitable term concepts for the focused contexts, with Precision = 0.7756, Recall = 0.7756 and F-Measure = 0.7735. Therefore, it can be concluded that the Technique of Semantic Unambiguity for a Concept Selection of Terms in Focused Contexts has high accuracy when applying the Reinforcement Learning.