2
$\begingroup$

Looking for a system which can generate answers to questions. Most systems and blogs posted on internet are on Question to answer but not on answer to question or paraphrasing or keyword to questions.
Seq2Seq I tried and even after training for many hours the results were not making sense.
Rule bases and template based systems like add What, who where etc to keywords have so many pitfalls. But if any system known giving decent outputs may also work. Kindly let me know if there is any such system known. Or if there is any other idea someone has then please suggest. I had above two ideas.

$\endgroup$
0
$\begingroup$

Question answering (QA) is a complex problem and an active field of research. There are probably some academic prototypes around, but I doubt there's any general-purpose ready-to-use QA library. However there are probably state of the art implementations for closed QA, i.e. QA restricted to a specific domain (I'm not aware of any specific library though).

Paraphrasing is a related but different problem, and also an active research question.

Extracting keywords is a much more standard task and is an important part of traditional Information Retrieval methods.

| improve this answer | |
$\endgroup$
  • $\begingroup$ What's wrong with my answer? $\endgroup$ – Erwan Dec 22 '19 at 0:03
  • $\begingroup$ Hi. You implied that it is not possible or you dont know. Now I doubt people who may know the answer may even answer this as my question has got one answer, but the answerer says he does not know. $\endgroup$ – Sandeep Bhutani Dec 24 '19 at 6:52
  • $\begingroup$ @SandeepBhutani your question is very general and mentions several very different NLP problems. You also seem to assume that the general problem of QA is easy to solve. So the point of my answer was: (1) to explain to you the differences between these problems and give you directions to each of them (the wikipedia pages give you links to the main papers or books about these topics). (2) to let you know that some of these problems are very complex, so they are not solved yet (and may never be solved). Of course NLP research evolves every day so I might not know the most recent... $\endgroup$ – Erwan Dec 24 '19 at 10:56
  • $\begingroup$ ... developments, but I've been in this field for the past 15 years and my guess is that it's not going to be solved anytime soon. If anybody has better information they can easily write an answer, but it might be useful to you or anybody else to understand what is feasible nowadays and what is not. $\endgroup$ – Erwan Dec 24 '19 at 11:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.