Menu
curious if anyone has insight into what algorithm google news uses to group like stories together? k-means? or something custom?
588
Algorithms personalize your news for these sections: For you. Topics, sources, and locations in Favorites. All other stories and notifications, except when stories are chosen by people.
Our articles and multimedia content are selected and ranked by computers that evaluate, among other factors, how often and on what websites a story appears online. We also rank based on certain characteristics of news content such as freshness, location, relevance, and diversity.”
Full Coverage organizes articles into storylines as the news event unfolds. There are no human editors involved in curating the stories and the results included in the Full Coverage section are not personalized. Everyone sees the same storyline.
“Top stories” is a section that appears within Google Search when we detect a search query is news-oriented. We match the search with relevant, quality news content to display. "Top stories" features articles related to the search, and a link to more related articles on the News tab.
It is kind of difficult to find that out, I guess; but for now I found this good white paper on possible algorithms for Google News Personalisation suggestions. Have a look for yourself:
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.80.4329&rep=rep1&type=pdf
The three algorithms covered here are: (1) MinHash clustering (2) Probabilistic Latent Semantic Indexing (3) Covisitation
and some combinations.
Hope this information was helpful!
52
When Google launched Google News, they used to put a small section about the algorithms they used to group on "About Google News" page, there was a mention of "An advanced Bayesian Network" and some other algorithms(no other algorithms names were mentioned!). That paragraph is now absent from the same page.
36
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With
