Project Strolling bot

FB Messenger chatbot that recommends users cultural events in Paris (exhibition, movie...) based on the user's interest.

Website: https://www.strolling.club/

Back-end Structure:

├── backend
│ ├── cinema
│ ├── exhibition
│ │ ├── expo_scraper
│ │ │ ├── extracted data
│ │ │ ├── expo_classifier
│ │ │ └── expo_scraper
│ │ │ └── spiders
│ │ └── README.txt
│ ├── language
│ ├── messenger
│ └── others
├── static
├── README.md
├── server.py
├── uwsgi_config.ini
└── wsgi.py

Environment

(virtualenv): Python 3.5

Installed libraries:

• flask

• uwsgi (not needed on local machines to dev)

• schedule -> to schedule python jobs

• scrapy -> to scrape website info

• BS4 -> to scrape trickier parts

• dateparser -> to transform dates in French in date objects

• nltk -> to work on text

• unidecode -> to get rid of accents, etc.

• pandas -> to concatenate text from different sources

• numpy -> to handle some table operations

• sklearn -> to use machine learning algorithms

• PIL (pillow) -> to transform image size, download version 4.3.0 (type in console: pip install 'pillow==4.3.0')

Sources de données

Besoin de rafraichir manuellement les données pour l'instant:

• expositions: aller dans /backend/exhibition/expo_scraper et run handle_exhibition_data.py
• films : aller dans /backend/cinema/ et run handle_cinema.py