Hi,
I need build a small application about Speech Recognition based-on Android platform, such as: port speech input to text.
Hence, is there a way/open source that to do for this?

Speech recognition in Linux - Wikipedia
http://en.wikipedia.org/wiki/Speech_recognition_in_Linux

List of speech recognition software - Wikipedia
http://en.wikipedia.org/wiki/List_of_speech_recognition_software

Category:Speech recognition software - Wikipedia
http://en.wikipedia.org/wiki/Category:Speech_recognition_software

Hello, Khanh Le, and welcome to chatbots.org!

I’m thinking that what you’re referring to is speech recognition for Android-based devices, such as smart-phones? If this is, indeed, the case, you may wish to check out the following links:

http://appinventor.googlelabs.com/about/ - Google App Inventor

http://developer.android.com/index.html - Android Developer Community

http://developer.android.com/guide/index.html - Android Developer’s Guide

I hope this helps.

Thank your reply, Dave.
I’m also interested in chatbots. I would like to build a simple chatbot for Android based devices.

How I should start to create simple chatbot?

While I’m not familiar with the Android OS, from what I understand, the SDK for the Google App Inventor is JAVA based, so it may be possible to “port” the JAVA-based AIML interpreter Program D over to Android. I suspect, however, that it’s not going to be quite that easy. My suggestion at this point is to use the links I gave above to pose that same question to the Android Community, since they’re far more knowledgeable about that OS. As far as creating a “simple” chatbot, I can also direct you to several good tutorials on that subject, but I think it may be better for you to gain some familiarity with the Android OS first. At least, to my thinking, that would be the more efficient order of things.

I suggest KISS.  Find Eliza and use that as a model.  Don’t fight the threads for listeners, etc. for the server based interpreter Program D.  There may be much in it that will directly port into Android.  Still it uses too much memory for Android apps.

The speech-to-text and text-to-speech in Android is only a small programming challenge - nothing to be concerned about I’m guessing.  I think it is done on the Google servers, so you’ll need internet access in your app.

Someday I might port my C# AIML interpreter (http://AIMLpad/ChatBot.zip) to Android using the sqllite for accessing the categories.  That way I don’t have to deal with big files of XML loading at start up.

I’m thinking one activity and one data provider service is all that is required unless the speech-to-text needs some special handling because of internet lags/delays.

Or you could create a webview activity and leverage Pandorabot - even simpler!  Check out a javascript bridge in a page you have included in the resources of your app.  I used something like that to position a Google map page using the phone’s GSP once.