[WEEK I–BOOK GENRE PREDICTOR]
Theme: Genre prediction from book summaries using Natural Language Processing and neural networks.
Team Members: Hakan AKYÜREK, Sefa YURTSEVEN
About the Project
We are going to use deep learning methods to achieve our goals in the project and if we have enough time we are thinking of adding extra features as well, this is why we are choosing deep learning over naive bayes. We have not came to a conclusion about the methods we are going to use just yet.
Our algorithm is going to predict a book’s genre from it’s summary. There are a vast number of books with various genres in our dataset. We shouldn’t have a problem with data gathering.
The project plan can be summarized in three steps:
(1) Cleaning the dataset
(2) Predicting genres from book summaries
(3) Adding extra features like predicting the author of a book if everything goes on course
We should note this is the inital plan and might go under some changes during development and further planning.
DATASET
The dataset we found contains around 16000 book summaries along with their genres, publication dates, authors and other informations.
Related Works
Here is a master thesis about text classification. The author of the thesis used score comparison method, we are planning to use neural networks, but still it contains useful information for us about other various problems.
We will of course probably find more related works
