A project we made for HACKYEAH 2019. (Team of 2) (Unfortunately we couldn't submit a working demo by the deadline as we encountered a database related bug at the last minute 😔 )
We created a rating system for articles and news sources using Natural Language Processing techniques to provide a more insightful browsing experience.
I aim to develop this further and create a rating system that becomes an industry standard. The goal is the create an emoticon like badge system that shows besides every article and news sources on Twitter,Facebook,Google and other apps where people consume media.
Right now the app is not very accurate as we have only allowed it to use 4 sources (bbc-co-uk, abc-news, the-washington-post, associated-press) and only provided it with a handful set of articles.
List of technologies we used: *Django for UI *Materialize CSS *Firestore for Database *PKE for keyword extraction *nltk for tokenizing *textBlob for sentiment analysis *BeautifulSoup for web scraping *NewsApi as a search engine for finding articles *Trello for project management