Integrated Question Classification based on Rules and Pattern Matching
Sanjay Kumar Dwivedi, Vaishali Singh · 2014
In question answering, task of question classification has remained in sharp focus since long. This paper presents our research about question classification in higher education domain. The questions for the domain are divided into 9 coarse grained categories and 61 fine grained categories. In the proposed integrated classification method, rules are adopted to extract coarse grain categories and focus words while pattern matching is implemented to classify the questions in fine grained categories. Experimental result shows the efficiency of our method which achieves satisfactory performance with the state of the art in this field.