Auralist: introducing serendipity into music recommendation

@inproceedings{Zhang2012AuralistIS,
  title={Auralist: introducing serendipity into music recommendation},
  author={Yuan Cao Zhang and Diarmuid {\'O} S{\'e}aghdha and Daniele Quercia and Tamas Jambor},
  booktitle={WSDM '12},
  year={2012}
}
Recommendation systems exist to help users discover content in a large body of items. [] Key Method Using a collection of novel algorithms inspired by principles of "serendipitous discovery", we demonstrate a method of successfully injecting serendipity, novelty and diversity into recommendations whilst limiting the impact on accuracy. We evaluate Auralist quantitatively over a broad set of metrics and, with a user study on music recommendation, show that Auralist's emphasis on serendipity indeed improves…

Figures and Tables from this paper

Music recommenders based on hybrid techniques and serendipity
TLDR
TRECS, the live prototype system, presents a weighted hybrid recommender approach that amalgamates three diverse recommender techniques into one comprehensive score and peppers the generated result list by some simple serendipity heuristic, so users can benefit from recommendations aligned with their current taste in music while gaining some exploratory diversification.
A Serendipity Model for News Recommendation
TLDR
This work proposes a content-based recommendation technique with the focus on serendipity of news recommendations, and proposes a general framework that incorporates the benefits of serendipsity- and similarity- based recommendation techniques.
A model for serendipitous music retrieval
TLDR
A new model is presented that combines several factors important for personalizing retrieval results: similarity, diversity, popularity, hotness, recentness, novelty, and serendipity, and the use of social media mining techniques to address the problem of estimating popularity and hotness in a geo-aware manner.
Context-Aware Music Recommendation with Serendipity Using Semantic Relations
TLDR
This system proposes a way of finding suitable but novel music according to the users' contexts and incorporates concept of serendipity using 'renso' alignments over Linked Data to satisfy the Users' music playing needs.
SIRUP: Serendipity In Recommendations via User Perceptions
TLDR
The model, called SIRUP, is inspired by the Silvia's curiosity theory and aims at measuring the novelty of an item with respect to the user profile, and assessing whether the user is able to manage such level of novelty (coping potential).
Tailoring Music Recommendations to Users by Considering Diversity, Mainstreaminess, and Novelty
TLDR
A comprehensive evaluation of a variety of music recommendation approaches (stand-alone and hybrids) on a collection of almost 200 million listening events gathered from Last.fm highlights cases where the proposed diversity, mainstreaminess, and novelty features can be beneficially integrated into music recommender systems.
HAES: A New Hybrid Approach for Movie Recommendation with Elastic Serendipity
TLDR
A new model called HAES is introduced, a new approach for movie recommendation with a more objective definition of serendipity, and a novel concept of elasticity in the recommendation, to adjust the level of serentipity flexibly and reach a trade-off between accuracy and serendipsity.
Escape the bubble: guided exploration of music preferences for serendipity and novelty
TLDR
This work presents a methodology and related system that allows users to initiate explorations of music genres by taking a gradual path towards the desired genre while viewing the preferences of other users.
Evaluating Hybrid Music Recommender Systems
TLDR
TRecS, the live prototype system, presents a weighted hybrid recommender approach that amalgamates three diverse recommender techniques into one comprehensive score and peppers the generated result list with recommendations based on a simple serendipity heuristic.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 32 REFERENCES
Proposal and Evaluation of Serendipitous Recommendation Method Using General Unexpectedness
TLDR
A human preference model of serendipitous items based on actual data concerning a user's impression collected by questionnaires was devised and evaluation results show that one of these recommendation methods, the one using general unexpectedness independent of user profiles, can recommend the serendippitous items accurately.
Metrics for Evaluating the Serendipity of Recommendation Lists
In this paper we propose metrics unexpectedness and unexpectedness_r for measuring the serendipity of recommendation lists produced by recommender systems. Recommender systems have been evaluated in
From hits to niches?: or how popular artists can bias music recommendation and discovery
TLDR
The results from the experiments reveal that---as expected by its inherent social component---the collaborative filtering approach is prone to popularity bias, which has some consequences on the discovery ratio as well as in the navigation through the Long Tail.
A new approach to evaluating novel recommendations
TLDR
A user-centric experiment shows that even though a social-based approach recommends less novel items than the authors' CB, users' perceived quality is better than those recommended by a pure CB method.
Improving recommendation lists through topic diversification
TLDR
This work presents topic diversification, a novel method designed to balance and diversify personalized recommendation lists in order to reflect the user's complete spectrum of interests, and introduces the intra-list similarity metric to assess the topical diversity of recommendation lists.
Taxonomy-driven computation of product recommendations
TLDR
Relationships between super-concepts and sub- Concepts constitute an important cornerstone of the novel approach, providing powerful inference opportunities for profile generation based upon the classification of products that customers have chosen.
Optimizing multiple objectives in collaborative filtering
TLDR
A general recommendation optimization framework that not only considers the predicted preference scores but also deals with additional operational or resource related recommendation goals and demonstrates through realistic examples how to expand existing rating prediction algorithms by biasing the recommendation depending on other external factors such as the availability, profitability or usefulness of an item.
Music Recommendation and the Long Tail
TLDR
A new service explicitly designed to make recommendations from the long tail is described, and popularity effects across the recommendations which it suggests are analyzed.
A Survey of Collaborative Filtering Techniques
TLDR
From basic techniques to the state-of-the-art, this paper attempts to present a comprehensive survey for CF techniques, which can be served as a roadmap for research and practice in this area.
Beyond Algorithms: An HCI Perspective on Recommender Systems
TLDR
From a user’s perspective, an effective recommender system inspires trust in the system; has system logic that is at least somewhat transparent; points users towards new, not-yet-experienced items; provides details about recommended items, including pictures and community ratings; and finally, provides ways to refine recommendations by including or excluding particular genres.
...
1
2
3
4
...