Rank and relevance in novelty and diversity metrics for recommender systems

By Saul Vargas, Pablo Castells. ACM Conference on Recommender Systems, 2011

755 citations recorded by the source corpus.

Subfields: Information Systems, Management Science and Operations Research, Marketing

View Paper: https://www.semanticscholar.org/paper/9fdb13b1da665e3041d725a8d5ef62b78861c0ea

How researchers see this paper

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

In Information Systems

In Management Science and Operations Research

In Marketing

Citations in the corpus

A historical perspective of explainable Artificial Intelligence (2020, background): “6 Here, novelty and diversity are concepts that relate to (serendipitous) information discovery which have been studied, among others, in the fields of information search and recommender systems (Clarke et al., 2008; Vargas & Castells, 2011).”
Exposure diversity as a design principle for recommender systems (2018, background): “As a way to mitigate this effect and enhance user satisfaction, recent approaches to recommender systems have started investigating and incorporating diversity into their design (Lathia, Hailes, Capra, & Amatriain, 2010; Ozturk & Han, 2014; Vargas & Castells, 2011).”
News recommender systems - Survey and roads ahead (2018, neutral): “An alternative definition was provided by Vargas and Castells [172], who defined novelty as the inverse general popularity of an item.”
Recommender systems survey (2013, background): “To evaluate these aspects, various metrics have been proposed to measure recommendation novelty and diversity [105,220].”
Extreme Multi-label Loss Functions for Recommendation, Tagging, Ranking & Other Missing Label Applications (2016, methodology): “Recommending rare items: It is important to remove the popularity bias and recommend rare/novel items in many applications [8, 11, 34, 38, 43, 46, 48, 50].”
Diversity, Serendipity, Novelty, and Coverage (2016, methodology): “Zhang [2013] identified three qualities of a novel recommendation: being unknown to the user, being relevant to the user, and being dissimilar to items in the user’s profile.”
The Effect of Recommendations on Network Structure (2016, background): “There is a variety of work that attempts to measure and address such concerns in network recommendations [26, 29].”
Multistakeholder recommendation: Survey and research directions (2020, background): “e is a large body of recent work on incorporating diversity, novelty, long tail promotion and other metrics as 10 additional objectives for recommendation generation and evaluation. See, for example, [1, 126, 146, 133, 69]. There is also a growing body of work on combining multiple objectives using ”
Evaluating Recommender Systems with User Experiments (2015, background): “A steady stream of research has considered algorithmic solutions to diversifying recommendations [1, 76, 115, 124, 125].”
Auralist: introducing serendipity into music recommendation (2012, background): “[25] S. Vargas and P. Castells.”
A probabilistic model to resolve diversity–accuracy challenge of recommendation systems (2015, background): “Self-information-based novelty is a measure for novelty relative to popularity of items [33,34].”
Current challenges and visions in music recommender systems research (2017, background): “Common choices are inverse cosine similarity [150], inverse Pearson correlation [183], or Hamming distance [101].”
Deep Learning Based Recommender System (2017, background): “ven lead to over-specialization. Apart from accuracy, other evaluation metrics such as, diversity [3, 102], novelty, serendipity, coverage, trustworthiness, privacy, interpretability etc. also ma−ers [29, 50, 76, 110]. For example, recommender system will bring customers with unclear and uncertain i”