Novelty and redundancy detection in adaptive filtering

By Yi Zhang, Jamie Callan, Tom Minka. Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, 2002

511 citations recorded by the source corpus.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/2f7f81529c67e9d98722f4ee17fe3808f0752b5e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Collaborative filtering recommender systems (2013, background): “Evidence demonstrates that customers tend to rely more on recommendations from their friends than on recommendations from similar but anonymous users (Zhang et al ., 2002).”
Information Retrieval Evaluation (2011, background): “For some discussion of the problems, see [81, 203].”
Cluster-based retrieval using language models (2004, methodology): “A different model for mixing these three sources has been pr oposed for novelty detection in adaptive filtering [28].”
A Distributed, Architecture-Centric Approach to Computing Accurate Recommendations from Very Large and Sparse Datasets (2011, background): “Zhang, Callan, and Minka argue that documents created at or around the same time are more likely to contain redundant information [173].”
Research-paper recommender systems: a literature survey (2015, background): “Other factors that affect user satisfaction are confidence in a recommender system [264], data security [345], diversity [346], user tasks [87], item’s lifespan [347] and novelty [348], risk of accepting recommendations [349], robustness against spam and fraud [350], transparency and explanations [3”
Content-based Recommender Systems: State of the Art and Trends (2011, background): “[102] to evaluate whether a document that is deemed to be relevant contains some novel information as well.”
A Comprehensive Survey of Neighborhood-based Recommendation Methods (2011, background): “Solutions proposed for this problem include adding some randomness [71] or filtering out items that are too similar [8, 77].”
Retrieval and novelty detection at the sentence level (2003, methodology): “Although cosine distance has performed very well as a novelty measure in past research that assigned novelty scores to full documents [14], its performance is known to degrade substantially on shorter pieces of text.”
Search result diversification (2010, background): “Another work based on novelty is [19], which aims at enhancing adaptive filtering systems with the capability of distinguishing novel and redundant items.”
Recommender Systems (2016, unknown): “The work in [670] showed that serendipity can be achieved in a recommender system with the help of user labeling.”
Learning to model relatedness for news recommendation (2011, methodology): “Relevance and novelty measures have been well studied [28, 33, 24, 39, 3], but it is not clear how well they would work in a unified relatedness function.”