- A new ranking algorithm to rank the best news articles. The old one was based on votes only - that is, the top ranking news item was the one with most votes. The new one is based on number of votes and time. So a new vote on a newly submitted news item will be worth more than a new vote on an old news item. If you want to learn more about ranking algorithms, you can visit: Reddit, Stumbleupon, Del.icio.us and Hacker News Algorithms Exposed! We have also saved the old algorithm, if you want to find the best articles (the one with most votes - not adjusted for time) from this week, month, year, or beginning of time. You will find them in the boxes labelled "Best News This..."
- A slightly improved design
- Related articles to one article. If you click on the comments section, you will find a box labelled "Related articles". The users who liked the article you are looking at did also like the articles in the box
- Recommended articles you may like. If you click on your @name at the top - you will find your saved articles and also a box labelled "Recommended Articles". The links in that box are based on what you have already voted on and what similar users to you have voted on.
- Categories. We currently have 4 categories: Markets, Economics, Personal Finance, and Management. You can't choose the category yourself, the computer will chose the category automatically when you submit a new article and every article may not have a category. We may add or change some categories in the future, so if you have a suggestion on a new category, please contact us.
The algorithms for the related, and recommended articles, and the category chooser comes from the excellent book Programming Collective Intelligence by Toby Segaran who is working at Google. The code in the book was based on the Python programming language, and we have translated it to PHP to make it work on Trejdify. You can find the translated code here at Github if you need it to your own projects.
Go and check out the latest version of Trejdify