Hacker News Reader

Comment by vadimbaryshev | Hacker News Reader

Hello HN,

I created trending news aggregation service called TOP.ST. Ratings are based on people's social network activity and updates in real-time. At the moment the service works in 29 countries. More countries are coming in future. In next updates I planning to add news categories and thinking about e-mail digest.

I'm looking for a feedback. Feel free to ask me any questions here or by e-mail: mail@top.st

View on HN

It looks good and seems to capture the vibe of the moment well.

Any details on how it works? Which networks do you monitor? Is it just Twitter? How do you know where people are from? Do you go by language? Only take into account tweets with geo information?

Yes! I too would love to know how it works under the hood!

Checkout buzzsumo.com and a free version that we built wordlink.com. It gathers all the news in the world (mostly) then sorts it by its aggregated social activity. Likes shares comments tweets pins upvotes, on facebook twitter linkedin google plus reddit pinterest etc... It's also a search engine used by marketing companies to analyze the news. Scrapy twisted python crawlers and elasticsearch backend for indexing.

Example links: http://wordlink.com http://wordlink.com/video http://wordlink.com/barack+obama http://wordlink.com/nytimes.com

I monitoring 7 networks: Facebook, Twitter, Google Plus, LinkedIn for the World and VK, Odnoklassniki for Russia. GEO targeting by source popularity in country.

Not OP but I'll hazard a guess it looks for URLs which appear at an accelerating rate on Twitter, using a whitelist of domain sources (since we see several frequent domains).

One feedback is I'd avoid the animation effect until the user has been idle for a few seconds.

Excellent idea about waiting for idle to update ratings. I thought a lot about how to prevent changes while user scrolling the list. At this moment user have a pause button, to do it manually, but your suggestion looks more useful. Thanks!

Nice service! It would be interesting to see some grouping of stories. Some stories appear many times from different sources.

Yes, I agree it would be a great feature. But it looks like that I need some more knowledge in machine-learning, to do this job well. Maybe in future I will be able to do it.

This is very resourceful, you should make a dark theme that's easy on the eyes, that would make this perfect.

Thank you. I already have a dark concept and I will try to release it as soon as I find a good place to put a switch.

Reply on news.ycombinator.com