I have used Binance platform for crypto trading for a few months now and learned how hard it is to know whether the cryptocurrency prices would rise or dip, at any given time.
What it does
CNS (Crypto News Sentiment) shows the cumulative negative and positive sentiments scores in real-time. This is done by extracting any given string of text taken from news articles' headline and summary. Then, the given text will be passed on to the expert.ai API sentiment endpoint as part of the request body and finally, getting a response to be shown in the user interface. With the use of the sentiment API of expert.ai, it is made easier to extract the sentiments and come up with the score for a given cryptocurrency, at any time.
How we built it
For the front end, I've used Angular 12 while for the backend, I have used expert.ai sentiment API and RapidApi's Newscatcher endpoint. It is currently hosted in AWS S3 as a static web site with the following URL: http://crypto-news-sentiment-opinion-mining-nlp-api-2021.s3-website-ap-southeast-1.amazonaws.com/
Challenges we ran into
Setting up Angular and AWS took a while, but I was able to get help from my former colleagues at work.
Accomplishments that we're proud of
I've done the development at a very short period. But there will be more development work to come.
What we learned
The use of expert.ai API, Angular 12, and AWS S3.
What's next for Crypto News Sentiments
Due to lack of time and preparation, I've only made 40% of the core functionality. The most important features would allow the users to click on each panel to display actual sentiment scores and other metadata to supplement the info. Other cool features to add are dashboards, graphs, points systems, and tentatively having it hosted in a subdomain under softwarebeing.org