Currently I have been working with speech recognition in java, and more or less have tried out sphinx where in I have been able to successfully get the word the person is trying to speak. So functionally I am trying to automate a software where in a person will speak, and my code will do the required action in java.

But what approach would be the best to do so. I need to evolve my code so that my code will train as and when it encounters more data, but that is still long. Currently how should I move on after I am done recognising the data.

For eg: launch the browser, here launch and browser are two keywords and it means I will have to invoke a method to launch the browser.

Which approach would be the best fit for me. I was planning to use drools rule engine, but any other suggestion better than this would be most welcome.

  • 2
    Are you asking how to do actions such as launching a browser, or are you asking how to develop a grammar which is convenient for your users to speak?
    – Cort Ammon
    May 1, 2017 at 16:48
  • No considering I know how to do actions, but I just wanted an approach that how should I develop the logic of calling such methods doing actions after I have done recognising words.I know I can do it using if else logic, but what would a best approach?
    – log N
    May 1, 2017 at 17:05
  • My recommendation would be to do the if/else logic for now, until you have more use cases. There's perhaps a few dozen different ways to handle dispatching commands like this, each with their own pros and cons. For example, there's a big difference between what you want for "launch firefox" and "open Excel" and "tell Excel: set text of current cell to 'hello world'" (the latter is phrased in apple-script notation to fit well with how AppleScript dispatched commands)
    – Cort Ammon
    May 1, 2017 at 17:18


Your Answer

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

Browse other questions tagged or ask your own question.