The Comparison of Algorithms for Thai-Sentence Classification

Thanyarat Nomponkrang, Charun Sanrach
2016 International Journal of Information and Education Technology  
The development of an online discussion board in collaborative learning makes available of tracking behavior and collaboration among students. The classification of discussion sentences on the board is an essential mechanism used to describe students' interaction patterns. This paper proposes feature extraction module that is used to extract the feature of Thai-sentence according to the sentence function. The module extracts two main features which are term binary (TB) of key phrase and term
more » ... phrase and term frequency (TF) of part-of-speech (POS).The TB term is used to indicate the presence of key phrase that cannot identified by POS and the TF term is used to calculate the term frequency-inverse document frequency (TF-IDF).Each feature is extracted from Thai-sentence to perform the data setswhichare1) TF of POS 2) TB and TF of POS 3) TF-IDF of POS and 4) TB and TF-IDF of POS. The performance of all data set is compared using 4 classification algorithms including: Decision Tree, Naïve Baye, K-nearest neighbor (k-NN) and Support vector machine (SVM).In this experiment shows the result in two dimensions which are the appropriate algorithm and the appropriate features to classify the Thai-sentence. The result is SVM algorithm is the optimal model on the dataset that have key phrase and TF-IDF term of POS. 808 Charun Sanrach received the Ph.D. degree in diplôme de docteur (informatique) INPL, France. His research interests include applying artificial intelligence in computer education, applying technology for learning and teaching, problem-based learning, project-based learning, and constructivist. Currently, he is with the
doi:10.7763/ijiet.2016.v6.796 fatcat:glibaptjijaj7fbl4gfjjvju4a