Use in conjunction with Frontend and ML model.
BrandPerception is an NLP powered tool to assist marketing teams in understanding the market sentiments of their brand.
- Run node index.js to start the application.
- Connect to the API on port 3000.
| HTTP Verbs | Endpoints | Action |
|---|---|---|
| GET | To retrieve posts with keyword, iphone | |
| POST | /predict | To predict for 1 sentence |
| Query String | Description | Mandatory? | Example |
|---|---|---|---|
| q | Retrieves posts with these specific keywords | Y | iphone |
| subreddit | Retrieves posts in this subreddit with these specific keywords | N | apple |
| before | Retrieves posts before x (s,m,h,d) | N | 2h |
| after | Retrieves posts within the past x (s,m,h,d) | N | 4d |
| score | Retrieves ones with this score operator. May not be accurate | N | (=/>/<) 2 |
| size | Retrieves stated size. Default is 500 and it is the maximum. | N | 50 |
Example
Gets posts with "iphone" in subreddit "apple" between 2 and 4 days ago with score more than or equals to 2, limited to 10.
/reddit?subreddit=apple&q=iphone&before=2d&after=4d&score>=2&size=10
Output
Data in array with each of the object having
'body', 'score', 'author', 'permalink', 'created_utc'