Relevance ranking for one to three term queries
Information Processing and Management: an International Journal
NLDB '02 Proceedings of the 6th International Conference on Applications of Natural Language to Information Systems-Revised Papers
Question Answering from Frequently Asked Question Files: Experiences with the FAQ Finder System
Question Answering from Frequently Asked Question Files: Experiences with the FAQ Finder System
Analyses for elucidating current question answering technology
Natural Language Engineering
Question answering passage retrieval using dependency relations
Proceedings of the 28th annual international ACM SIGIR conference on Research and development in information retrieval
MultiSumQA '02 proceedings of the 2002 conference on multilingual summarization and question answering - Volume 19
Interesting nuggets and their impact on definitional question answering
SIGIR '07 Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval
Finding the right facts in the crowd: factoid question answering over social media
Proceedings of the 17th international conference on World Wide Web
Hi-index | 0.00 |
When a user finds the answer using question answering system, sometimes the user may not be satisfied with answers. This is because it is difficult to deliver the intention of a user question to the system due to the lack of expression ability and polysemy and etc. So, the system must provide additional ways to search the correct answers. In this paper, we propose a new method to recommend questions that can satisfy the purpose of user question with high probability. We define the supplementary similarity to calculate the similarity between sentences which have a few common terms. And we define the fuzzy relation product operator to find the questions that are recommended when user is not satisfied with the answers. User can select the question among the list of recommended questions and has more chance to find the answers that the user can be satisfied with. We experiment and show that the performance of the system gets high as the recommendation is repeated.