An automatic web based Question Answering (QA) system is a valuable tool for improving e-learning and education. Several approaches employ natural language processing technology to understand questions given in natural language text, which is incomplete and error-prone. In addition, instead of extracting exact answer, many approaches simply return hyperlinks to documents containing the answers, which is inconvenient for the students or learners. In this paper we develop technique to detect the type of a question, based on which the proper technique for extracting the answer is used. The system returns only blocks or phrases of data containing the answer rather than full documents. Therefore, we can highly improve the efficiency of Web QA systems for e-learning.
The goal of an intelligent answering system is that the system can respond to questions automatically. For developing such kind of system, it should be able to answer, and store these questions along with their answers. Our intelligent QA (iQA) system for Arabic language will be growing automatically when users ask new questions and the system will be accumulating these new question-answer pairs in its database. This will speed up the processing when the same question(even if it is in different syntactical structure but semantically same) is being asked again in the future. The source of knowledge of our system is the World Wide Web(WWW). The system can also understand and respond to more sophisticated questions that need a kind of temporal inference.
In this study, we propose a hybrid Question Answering (QA) system for Arabic language. The system combines textual and structured knowledge-Base (KB) data for question answering. It make use of other relevant text data, outside the KB, which could enrich the available information. The system consists of four modules. 1) a KB, 2) an online module, and 3) A Text- to-KB transformer to construct our own knowledge base from web texts. Using these modules, we can query two types of information sources: knowledge bases, and web text. Text-to-KB uses web search results to identify question topic entities, map question words to KB predicates, and enhance the features of the candidates obtained from the KB. The system scored f-measure of .495 when using KB. The system performed better with f-measure of .573 when using both KB and Text-to-KB module. The system demonstrates higher performance by combining knowledge base and text from external resources.
Question Answering (QA) Systems are systems that attempts to answer questions posed by human in natural language. As a part of the QA system comes the question processing module. The question processing module serves several tasks including question classification and focus identification. Question classification and focus identification play crucial role in Question Answering systems. This paper describes and evaluates the techniques we developed for answer type detection based on question classification and focus identification in Arabic Question Answering systems. Question classification helps in providing the type of the expected answer and hence directing the answer extraction module to apply the proper technique for extracting the answer. While focus identification helps in ranking the candidate answers. Consequently, that has increased the accuracy of answers produced by the QA system. Question processing module involves analysing the questions in order to extract the important information for identifying what is being asked and how to approach answering it, and this is one of the most important components of a QA system. Therefore, we propose methods for solving two main problems in question analysis, namely question classification and focus extraction.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.