7.2. Related work: novelty and diversity filtering

In Pub/Sub systems, users are often faced with an issue of notifications, which tend to be flooded by information. To enhance the quality of filtering, two properties of information were proposed: novelty and diversity. In filtering by novelty, the objective is to discard an item whose information has already been notified (truncate or a similar content). Diversity captures a complementary kind of redundancy since it measures whether the information contained in a given item is globally present in the set of recently notified items or not. We present here how these two properties are used in the literature, first on the Web and then in a Pub/Sub context.

When searching a document on the Web, we assume static and already known documents. The objective is to present to users the k most relevant and diverse documents matching a query. To achieve this, some models are based on probabilities for diversity [ANG 11] or on graphs for computing...