Unsupervised Latent Dirichlet Allocation for supervised question classification     

Saeedeh Momtazi 《Information processing & management》2018,54(3):380-393

Question answering systems assist users in satisfying their information needs more precisely by providing focused responses to their questions. Among the various systems developed for such a purpose, community-based question answering has recently received researchers’ attention due to the large amount of user-generated questions and answers in social question-and-answer platforms. Reusing such data sources requires an accurate information retrieval component enhanced by a question classifier. The question classification gives the system the possibility to have information about question categories to focus on questions and answers from relevant categories to the input question. In this paper, we propose a new method based on unsupervised Latent Dirichlet Allocation for classifying questions in community-based question answering. Our method first uses unsupervised topic modeling to extract topics from a large amount of unlabeled data. The learned topics are then used in the training phase to find their association with the available category labels in the training data. The category mixture of topics is finally used to predict the label of unseen data.  相似文献   

Viability of construct validity of the speaking modules of international language examinations (IELTS vs. TOEFL iBT): evidence from Iranian test-takers     

Keivan Zahedi  Saeedeh Shamsaee 《Educational Assessment, Evaluation and Accountability》2012,24(3):263-277

The aim of the present research is to examine the viability of the construct validity of the speaking modules of two internationally recognized language proficiency examinations, namely IELTS and TOEFL iBT. High-stake standardized tests play a crucial and decisive role in determining the future academic life of many people. Overall obtained scores of candidates are believed to reflect their general proficiency level. Appropriate interpretation and use of test scores depend on the extent to which items measuring a particular skill (here speaking) can meet the criteria to examine the intended construct. Speaking, amongst the other four skills, has a central place in assessing general proficiency of the candidates. This research seeks to scrutinize how IELTS and TOEFL iBT tap on the speaking proficiency of their candidates. Moreover, this study investigates whether obtained speaking scores of candidates in these two international high-stake tests show an acceptable degree of consistency in measuring the skill being examined. The chosen sample of the study consisted of 60 students who successfully completed TOEFL iBT and IELTS preparation courses in Tehran. The results of the statistical analysis show that there is a meaningful discrepancy between the two exams in assessing the speaking abilities of the exam-takers and therefore challenge the construct validity of the exams in question. Findings are then used to discuss the repercussions for language proficiency measurement and assessment.  相似文献