Real-time processing and natural language processing are one of the most sought after technologies in the analytics stack, needed by organizations today. This application showcases these techniques to the best. Extracting latest trending news stories from all over the internet, streaming it to the machine learning algorithm which summarizes and finds important points is the prime ability of this application.
True News showcases the summarized text in an infographic format which can be give a quick overview without diving into too many details. The end result is a set of beautiful picture posts from all over the internet. Moreover, users can type a search string and get picture posts from all over the internet. The end to end flow is as below:
- Pull data from news API’s using Kafka.
- Cleanse and optimize it using Spark Streaming
- Summarize it and Predict if Fake using Spark ML Lib, PMML libraries.
- Display it using Infographic.
Additional features of the text summarizer into app include:
- Assign influencers rank by verifications, views, comments, reactions, positive sentiments over time, etc
- Filter one reaction/comment per article per user per IP address.
- Show only top commented articles from influencers.
- Show both sides of the views equally.
- Ban users activity if found guilty of having multiple comments, ids, articles from the same IP.
Watch a sample MVP demo here.