KILLER HUNTERS - Solving a murder mystery

Inspiration

• Though fictional, everyone would have at some point of time during childhood dreamt of becoming Sherlock Holmes. To gain that remarkable power of deduction and acute mental abilities. As a group of friends we find ourselves indulged in playing the boardgame called, Clue. From the description of each cast of the game to the weapons used, everything excites us (and million others! :).

• This project is a tactic to enjoy the overall experience of learning (by incorporating the above two) and working on this new and exciting language (prolog).

What it does

• We came up with two different models to find the culprit.

• One way is by matching the clues and evidences that were collected from crime scene. Second and more interesting way is based on checking consistency of suspects’ testimonies. Both the models will return the probable culprit based on the facts and rules defined.

How we built it

• We devised a murder spot with evidences and how they are related to suspects. We worked on evidence driven approach first. Later, we also wanted to try the testimony approach and successfully completed it.

Challenges we ran into

• Prof. Gopal Gupta's workshop was really an eye-opener as everything we heard was first off. Though it was intriguing to do the logical programming. Being used to code for years in python & Java, Prolog demanded a complete new perspective with the way we approach devising the problem solution.

• Apart from learning new syntaxes, we faced difficulties in validating testimonies as the argument had 6 possible values. We had to make sure that there are no logical loopholes in the rules as there were multiple evidences/identities in the scene.

• We researched a lot in performing List operations as List is the building block for our testimony approach. This also enhanced our understanding about the language and its application.

• The mentors were always there to clarify and direct us towards the right sources.

Accomplishments that we're proud of

• Got a fair idea on prolog programming.
• Deriving conclusion by inferring from inconclusive evidences.
• Learnt how to make use of rules and facts to find solution, how to conjunct rules, making use of negation as failure and so on.

What's next for Killer Hunters

• We would like to enhance our engine to also consider
• behavioral aspects of the suspects (during the interrogation)
• rate each individual on their motive to kill the victim from the past incidents for more assurance on the solution.