[Corpora-List] need a help

Eric Atwell E.S.Atwell at leeds.ac.uk
Wed Apr 12 16:25:41 CEST 2017


I *think* you want to match a Question "sentence" against possible Answer "sentences" in a text document, assuming the best match is the correct Answer.

I am just back from EACL'2017 European conference of the Association for Computational Linguistics, in Valencia Spain; and semantic analysis and similarity-matching of text was a "hot topic"; for example:

E Nagoudi and D Shwab "Semantic Similarity of Arabic Sentences with Word Embeddings". Proc WANLP at EACL'2017 http://aclweb.org/anthology/W17-1303

This is also the focus of annual SEMEVAL workshops, eg see SemEval-2017 http://alt.qcri.org/semeval2017/index.php?id=tasks

... and for data mining enthusiasts, Kaggle.com is a social media site for sharing data and data-mining methods; Kaggle hosts contests, and an ongoing contest (with $25,000 prize!) is "Quora Question Pairs: Can you identify question pairs that have the same intent?" https://www.kaggle.com/c/quora-question-pairs

- browse the Kaggle discussion for lots of methods you could try

I hope this helps


Eric Atwell, Associate Prof, Artificial Intelligence and Language at Leeds,

Best Lecturer 2017 award from COMPSOC Computing Student Society,

School of Computing, Univ of Leeds, Times University of the Year 2017


On Tue, 11 Apr 2017, Noura Labidi wrote:

> Hi all,
> I don't know if I could ask such a question;
> In fact I wanted to know if there are known approaches, methods and / or
> techniques for searching or detecting a response to a question in a text
> (the answer to the question is in a textual document).
> Cordially.
> Noura LABIDI.

More information about the Corpora mailing list