Abstract:
Movies are one of the most prominent means of entertainment. The widespread
use ot the Internet in recent times has led to large volumes of data related to
movies being generated and shared online. We can understand the importance
ot the movie as Netflix and Amazon Prime is one ofthe widely used sites for video
streaming and is earning a lot. Moreover, movies have now become one ofthe main
sources of entertainment for people. The extensive use of the Internet
has increased the creation and sharing of movie related data online. Movie plot
summaries generally tell about the movie genres and many people read them
before deciding to watch a movie. An automatic system can be applied to predict
genres based on summaries.
Movie plot summaries are expected to reflect the genre of movies since many
spectators read the plot summaries before deciding to watch a movie. In this study,
we perform movie genre classifications from plot summaries of movies using
machine learning model known as LSTM. We have used dataset from official
KAGGLE site named as “Kaggle Wikipedia movie plot” and have worked on the genre
and summary feature. Our model consists ofKeras embedding layer LSTM layer and
dense layer. The Project is broken down into multiple stages like data cleaning, pre processing, feature extraction, genre classification, training and testing and lastly
frontend and after doing our research we establish that according to our dataset comedy
and drama genre has higher accuracy then action and horror. Action has higher
precision 91%, drama has higher recall 92% and drama also has highest fl score of
72%
