SnappyFingers calls itself a comprehensive question/answer explorer. What this means is they index millions of FAQs spread across the web to give you a one-stop-shop for when you turn to the Web to have your questions answered.

We have interviewed the founder of SnappyFingers, Chirayu Patel to learn more about how SnappyFingers works and to get a glimpse behind the scenes of an answer search service in the making.

Pandia: How do you identify FAQs?

Chirayu Patel: Our primary source of FAQs has been Alexa. Alexa’s rich API makes it easy to discover high quality FAQs. We are also experimenting with Yahoo! BOSS.

Pandia: Do you index answers for every kind of question?

Chirayu Patel: SnappyFingers indexes questions and answers that can be found in FAQs spread across the Internet. Not all the questions in the FAQ are indexed though. Our relevancy algorithms filter out the unimportant, spam and duplicate questions.

Pandia: You have a link to “Similar questions” in your search results. How do you categorize FAQs?

Chirayu Patel: When a user views a question, we provide him with a set of similar questions that could provide more information. For a question “What is iPhone?”, we also display questions like “What is the resolution of the iPhone screen?”, “What is the battery life of the iPhone”.

Our current algorithm uses keyword matching to derive similar questions. Though it works quite well, there is still scope for improvement.

Pandia: How do you rank answers for your search results?

Chirayu Patel: Our relevancy algorithms assign a score to every question and answer based on a number of factors. This score not only decides the indexing eligibility, but also the ranking in search results.

Pandia: What are your plans for SnappyFingers?

We are working to further improve search results. We will do this by increasing the size of our database, and improving our relevancy algorithm. In the coming days, we plan to aggressively crawl the web, and roll out new versions of our relevancy algorithms.