2. Creating a Q-A system (Using Taming Text code)

During our research we came across a good book called “Taming Text”. This books 8th chapter is freely available to be seen/downloaded using this link: http://www.manning.com/ingersoll/Sample-ch08.pdf.
In this chapter, the author discusses how making a Q-A system using Lucene/Solr. Fortunately, he has given the source code of the Q-A system he explained in this book. I tried to follow the instructions given in the book and observed if this particular system can give us some insight on how to make the Q-A system of the likes of IBM Watson. After configuration (which took 10 minutes) and indexing (which took 5 Hrs), my Taming Text Q-A system was up and running. After getting the first hands on the system, I am not very satisfied with the results.

I am not sure if I have missed something, while configuration or indexing or some after steps. But, I am not getting any meaningful answers to my queries. For example, a simple query like “who was Muzzy Marcellino” gives pretty decent results.

But when I made this query a little bit complicated, the system didn’t give me any result.

Even though it would be unfair to compare this system with Google’s Q-A system, But our final goal is to create something which gives results which makes more sense.

Nevertheless, this exercise gave me insight on what things I should focus on, also a learning that it is not easy to create a generic Q-A system. For now I would be focusing on indexing information about the person type.