If you are like me who has 100s of articles saved to read for later and uses “Pocket” to browse offline, you might find it difficult to organize this list without tags. Pocket lets you add tags manually, but that means every time I add an article, I have to add tags and there is no way I am going to backfill those tags for a reading list of more than 100 articles.
Now Pocket does have a “suggested tags” service for their paid premium plans. This still requires manual work of adding the tags to each article one by one. So it doesn’t work for me since I always prefer automation over manual work.
In order to solve this problem, I wrote a python module that uses Google cloud’s Natural Language Processing API to automatically analyze the webpage from articles saved in my Pocket list, derives tags/keywords based on the content of the page, and adds relevant tags to the articles in my Pocket list for free.
Features
- Uses Python wrapper for Pocket API to retrieve articles in the My List
- Uses Beautiful Soup to scrape webpages
- Uses Google Cloud’s Natural Language Processing API to generate list of categories and entities from the content of the webpage
- Uses Pocket API to add tags to articles in the “My List”
Demo
Github Link
You can find more information about it and usage instructions on the github project page.