Google News: Story and Article Ranking
Ranking in Google News has two phases:
1. story ranking
2. individual article ranking
[edit] Story ranking
Story ranking appears on Google News as well as in real life, on the newspaper editors' desk. It's decided how important a story is. To illustrate this process with an example, a news story about President Obama's Nobel Prize is more important from editorial point of view than the appearance of a new hand cream on the market.
A story's rank is decided by "Aggregate Editorial Interest" and the process is very similar to the newspaper editors' story ranking protocol.
If a news event is picked up by a local newspaper and eventually a few news wires, this probably means that the editors world-wide think that the news event is not very important.
Yet, if a news event is picked up by many local newspapers, then many news wires and causes a global boom in publishing activity, it's sure that the story is important.
Example: if in a small town a grocer's shop is robbed, the story will be picked up by the local newspaper, maybe by some news wires as well. If Google decides to release its own operating system and its own PC, the story will have lot more interest for the newspapers as well as the news wires and will be picked up by many news providers globally. Even though both stories are news worthy and both of them will be included in Google News, the second story has higher "Aggregate Editorial Interest" thus the news cluster containing the story will be ranked higher.
[edit] Individual Article Ranking
When the stories were grouped in news clusters and the clusters' rank was decided, it's time to decide which individual article from within the cluster will get the most prominent 1st place, which article will get the 2nd place and so on.
There are many signals which Google News listens to when it decides individual articles' rank. To list some of the most interesting and important:
- Freshness, timeliness
- Duplication and novelty detection
- Local/personal relevancy
- Trusted source
1. Freshness - it's important to have the most recent, objective news. Having said "objective", Op-eds, satire, opinion articles can't occupy the first place in a cluster
2. Duplication and novelty detection - having original story is considered very important. Google News will attempt to detect which news source broke the story and will automatically assign the source higher rank
Example: If you break a story and other news sources rely on your reportage, you're in a winning situation
3. Local/personal relevancy - Google news gives more weight to sources which are local to the news event as it's considered their content is more relevant to these news events.
Example: if you're a publisher from London and you report a news event which happened in London, you will get higher rank than a publisher who reports the same news event but from New York
4. Trusted source - This is purely data driven. Google News observes, for example, how users interact with sources' news stories and based on these observations will assign higher rank for sources which users seem to like.
Example: if your article is placed on the 1st position of a news cluster but users choose to click on a headline which is situated on the 3rd position in the same cluster, that's a signal for Google News

