I'm working on a project,whereby; one of the actors is "Teacher" and it's role will be, insert the course name, outcomes of skills, then insert the question, student degree and automatically will be connected with appropriate skill through the natural language processing (NLP) analysis student degree.

What algorithm should I used in Natural Language Processing to connect the question to the appropriate skill (text comparison)?

  • 2
    $\begingroup$ Can you clarify your question? The wording is very confusing. Are there multiple questions? Where do the 'appropriate skills' come from? Can you provide an example? $\endgroup$ – Brian O'Donnell Oct 12 '18 at 13:55
  • $\begingroup$ How many skills are there? Do you have a fixed list of skills? $\endgroup$ – Martin Thoma Oct 13 '18 at 7:42
  • $\begingroup$ Do you already have labeled examples? Could you post some? $\endgroup$ – Martin Thoma Oct 13 '18 at 7:42
  • $\begingroup$ I like your question, because it is a common problem, that a question & answering system gets input as raw text and must extract from the data appropriate information by hand. In most cases, simple pattern matching algorithms won't work. Only specialized AI techniques which are utilizing deeplearning and agent architectures can solve these hard problems. Your question will motivate young researchers to think about the right algorithm more seriously. $\endgroup$ – Manuel Rodriguez Oct 30 '18 at 21:08

You can use the Bayes Text Classification algorithm which is widely used text classification. It works upon the Bayes theorem. You can know more from here.

  • $\begingroup$ The answer is a bit short; bayes classification is not completely wrong, but it must be explained in context. The better way is to introduce first the overall task, which is NLP The subproblem of the question is parsing, or to be more specific: stochastic parsing because it should be error tolerant. The first step is to create a grammar which includes items like coursename and skills. And this grammar is matched to the input string. The search term for further literature research would be “question parsing with stochastic grammar”. $\endgroup$ – Manuel Rodriguez Oct 30 '18 at 21:01

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.