To pique interest among young users to the fantasy genre, a media entertainment service provider has decided to publish intriguing plot summaries of some movies. You are hired as a Machine Learning specialist for this initiative.

Your task is to build an artificial intelligence-based model that generates plot summaries from scripts of these movies.

Data

Data files

Training data

• Scripts_Philosophers Stone.pdf: Script of Harry Potter & the Philosopher's stone [Source credits: https://www.scripts.com/]
• Plot summary_Philosophers Stone.txt: Sample plot summary

Test data

• Scripts_Goblet of Fire.pdf: Script of Harry Potter & the Goblet of Fire [Source credits: https://www.scripts.com/]

Submission format

• You are required to submit a .txt file containing the plot summary under 'Upload Prediction File'.
• You are required to upload your source code under 'Upload Source File'.

Evaluation criteria

Your text file submission will be evaluated using the following steps:

• Remove stopwords from the submission file.
• Data are tokenized to convert it to vector using MosesDetokenizer.
• Change sentences to vector using the TF-IDF Vectorizer.
• The vector for submission vs. actual is evaluated based on jaccard_distance $score = 100 * jaccard\_distance(submission, actual)$.