<![CDATA[Question Answering (QA) is a system that provide answer from question given from user. In QA there is one task called question analysis. Question analysis act as type chooser from query user input. Question analysis can be found with classification. This research using Naive Bayes as classification method. Furthermore, several process used from natural language processing such as question feature extraction and preprocessing contain data cleaning, stemming, stopword removal, and tokenization. Next phase is to build a classification model from training data which contain 16 question categories. Based on test result with 2 scenarios with preprocessing and without preprocessing, we obtained accuracy value of 0,58364 with preprocessing. We also obtained accuracy value of 0,65060 without preprocessing. Application of preprocessing in question classification have a negative impact because it changed the given question context.]]>
Copyrights © 2022
