Video algorithm question

Hi

Kindly help me with this question:

• A short design document detailing what you and why, which algorithms were used and what methods.
• Working code solving the problem, and sample output for the program.
• Please use Python for the implementation, unless you were asked otherwise. Please provide the complete source code. and an executable for compiled languages.

You are given two files, containing a list of movies, their plot summaries, and their genres. Each line of the file is a separate JSON object, representing one movie. Each movie object contains the fields: name, plot. genres and year. Using the file movies train.json as a training corpus, please create a program that takes arbitrary movie plots that do not appear in the training set, and guesses the genre(s) of the unknown movie, with a level of certainty. For example, given the plot "Five teens walk in to a creepy forest to be butchered by an evil psychopath" should output something like "Horror (85%), drama (20%)".
Use the file movies test.json to create a benchmark that evaluates the accuracy of your algorithm, including precision and recall measurements. The test and training json files are in the following links:
https://drive.google.com/file/d/0B1s...ew?usp=sharing
https://drive.google.com/file/d/0B1s...ew?usp=sharing

Thanks