Improvement of Morphological Analysis based on Three Layers of Lemmatization
Rabiah Abdul Kadir, Mary Ting, Hejab M. Alfawareh · 2024
Currently, the world impress with the ChatGPT. ChatGPT works with more than one Natural Language Processing (NLP) techniques to analyzed and inputted data in text form. Started from analysing the lexical and the syntactic of text, it is moved to pragmatic or discourse analysis. To process unstructured data such as natural language text, the apps should start with the morphological analysis. Morphological analysis plays a critical role in NLP to study the internal structure of words that involve with syntactic and semantic analysis. Lemmatization is one of the technique in morphological analysis involves with high-level processing techniques to increase the accuracy. This paper proposes a new model that incorporates the Stanford morphology class, WordNet lexical dictionary, adaptive learning framework, and longest-match algorithm in lemmatization process. The outcome of the proposed model shows tremendous improvement with 96% accuracy compared with other existing algorithms in morphological analysis. Improvement of the morphological analysis will increase the accuracy of the semantic meaning of any natural language text processing apps like chatbot, question answering system etc.