Shiraz: A Proposed List Wise Approach to Answer Validation

Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015), pages 220–225, Denver, Colorado, June 4-5, 2015. c⃝2015 Association for Computational Linguistics Shiraz: A Proposed List Wise Approach to Answer Validation Amin Heydari Alashty Saeed Rahmani Meysam Roostaee Mostafa Fakhrahmad Shiraz University Eram Street Shiraz, Iran heidari@cse.shirazu.ac.ir {srahmani,mroostaee,mfakhrahmad}@shirazu.ac.ir Abstract Answer Validation is an important step in Automatic Question Answering systems and nowadays by spreading Community Question Answering systems it is known as an impor- tant task by itself. Previous works just con- sidered it as a binary classification problem in which they try to find the best answer among all the candidate answers for a question. Ac- cordingly, they do not consider the possible unique information which may have been in- cluded other answers. This can be consid- ered by having a multiclass label classification problem, it is not only able to find the best answer but also can find ”potentially good”, ”bad”, and etc. answers too. By doing so, it is fully expected to extract and rate all the nec- essary information from existing candidates to help questioner to find the best and general an- swer for his question. This work tries to con- sider some features which are gained from im- portance of comments of the questioner. Fi- nally, by using a good classifier, we try to overcome this problem. The designed system participated in subtask A of the Semeval-2015 Task 3. The primary submission ranked at the 5th and 7th places in four class label and three class label evaluation, accordingly. 1 Introduction By spreading Community Question Answering (CQA) systems, there have been created a new tax- onomy for Question Answering (QA) systems: Reg- ular QAs, and CQAs. A regular QA, accepts a natu- ral language input and after searching into it’s avail- able resources, returns the best shortest answer, it could find. In these systems, answering to factoid questions may be an easier challenge than the other question types. One of the features of CQA systems is its users. Once one asks a question, others try to answer that question. Then these kinds of sys- tems just try to use users knowledge to answer users questions. Of course instead of finding the correct answer of an asked question from some candidate answers which must be done by questioner, system tries to tell the questioner which answer is helpful and which one is not. Then discussing about factoid questions is maybe so hard and it could not be han- dled just by using the answers and questions body, rather, it should have access to a great knowledge source to check if an answer is correct or not. Community Question Answering systems’ spreaded over the internet, and accordingly, it made researchers to be interested in getting involved to the challenges related to these systems. One of the main challenges which may be so important in the aspect of all the people who are using these systems is Answer Validation. More researches has been done as CQA systems …

Content truncated to 500 words