egemenzeytinci / foxandbox

Content based movie recommendation system

foxandbox

Installation and Usage

Deploy the smlar plug-in using the following commands,

$ git clone https://github.com/egemenzeytinci/smlar
$ cd smlar
$ USE_PGXS=1 make
$ USE_PGXS=1 make install

Then create smlar extension to PostgreSQL,

$ psql
test=# CREATE EXTENSION smlar;
CREATE EXTENSION

Install requirements,

$ pip3 install -r requirements.txt

You can create tables on PostgreSQL as follows,

$ python3 migration.py default.ini # or custom ini file

Then, you need to download the IMDB datasets to the temporary path under the system section in the config file,

$ wget https://datasets.imdbws.com/title.basics.tsv.gz -P /path/to/folder
$ wget https://datasets.imdbws.com/title.ratings.tsv.gz -P /path/to/folder
$ wget https://datasets.imdbws.com/title.episode.tsv.gz -P /path/to/folder

Then, please run DB seed script to insert static data,

$ python3 seed.py ./default.ini # or custom ini file

You can now run the crawler to fetch the movie images,

$ python3 crawl.py ./default.ini # or custom ini file

Please keep in mind, it takes a long time.

Then, please run the recommendation script to determine the clusters for the movies,

$ python3 recommendation.py ./default.ini # or custom ini file

You need to install Elasticsearch, then below commands will create indexes and save documents,

$ python3 indexing.py ./default.ini # or custom ini file

Finally, you can run web application as follows,

$ python3 app.py ./default.ini # or custom ini file