Misspelling Error Detection in Thai Language Application
Patsaraporn Somboonsak · 2018
This paper aims to create an application for verifying Thai grammatical errors. The cause of the problem is document inaccuracy. This development process started with dictionary construction for collection words and reference to Royal Institute of Thailand's principles. Next, using Thai word segmentation, which is composed of two methods by Thai Character Clusters (TCC), Statistical -- based method, Longest matching, Maximum matching, Left and Right Conditional Probability and Feature-Based Approach. Then calculate probability in word by high frequency and conjunctive in the database. Misspellings error detection for each character that is part of the words as no meaning or not in the dictionary is discarded. The experimental results show that our proposed application can provide detection and correction in average in formal documents and in informal documents. The investigation results of experiments using Thai document in both formal and informal with efficiency.