Question classification based on an extended class sequential rule model

Question classification is a crucial preprocessing for question answering system; it can help to make sure the user’s intention. Most of previous researches focus on the feature driven methods that represent a question with a bag of features, which ignore the important information contained in the words order and distance. To take such information into account, this paper proposes to describe the questions via the ExCSR (Extended Class Sequential Rule) model. To mine ExCSR rules, a method based on PrefixSpan, called DS-SRM (Distance Sensitive Sequential Rule Miner), is presented as well. Due to the existence of redundancy in the mined rules, a rule selection algorithm MCRSelection (Most Cover Rule Selection) is also proposed to find the most interesting rules. Experiments results on UIUC question set 1 show that the proposed method can achieve the accuracy of 90.6%, which outperforms the previously reported results.