Question posted in 
I am trying to develop an android (java) project for my Artificial Intelligent thesis. It is shortly based on story reading and word quiz. One person reads a story and marks the words that he doesnt know. These words are registered to WordPortfolio db that has “Word_id”, “Seen”(how many times), “Asked” (How many times asked in quiz), “Right”(how many times answered right).
I have “Words” table in my db that has 3 different parameters to make one word unique. Those are “Priority”, “Level” and a specifier whether it is a Verb, Noun, Adj, Adv. etc.
What I want to ask is;
Which algorithm can I use to classify these words to ask a “word-meaning question” wisely to the learner? I want learner to see the words he had seen in story-reading part more than one to consolidate the meaning of it and also I want him to learn new words.
2
Answers
There are many types of algorithms designed to do this. For instance, you could use a linear regression, nearest neightbor, clustering, or neural network. http://en.wikipedia.org/wiki/List_of_machine_learning_algorithms provides a pretty comprehinsive list of the options out there.
I would also see if your library has the book “Programming Collective Intelligence” by Toby Segaran (http://shop.oreilly.com/product/9780596529321.do) or something similar in your library.
classification and clustering algorithms has implemented in many artificial intelligence software’s such as MATLAB,WEKA,etc. you can see a sample of this in WEKA Text Classification for First Time & Beginner Users but i think your problem will have a good performance on MAP/REDUCE Freamework. I suggest you to use MAHOUT in your Problem which has a parallel framework and you can compromise your speed of other platforms with it.