CHESTNUT: Improve Serendipity in Movie Recommendation by an Information Theory-Based Collaborative Filtering Approach
19 Jul 2020-pp 78-95
TL;DR: This paper introduces CHESTNUT, a memory-based movie collaborative filtering system to improve serendipity performance and demonstrates a method of successfully injecting insight, unexpectedness and usefulness, which are key metrics for a more comprehensive understanding of serendipsity, into a practical serendIPitous recommender system.
Abstract: The term “serendipity” has been understood narrowly in the Recommender System Applying a user-centered approach, user-friendly serendipitous recommender systems are expected to be developed based on a good understanding of serendipity In this paper, we introduce CHESTNUT, a memory-based movie collaborative filtering system to improve serendipity performance Relying on a proposed Information Theory-based algorithm and previous study, we demonstrate a method of successfully injecting insight, unexpectedness and usefulness, which are key metrics for a more comprehensive understanding of serendipity, into a practical serendipitous recommender system With lightweight experiments, we have revealed a few runtime issues and further optimized the same We have evaluated CHESTNUT in both practicability and effectiveness, and the results show that it is fast, scalable and improves serendipity performance significantly, compared with mainstream memory-based collaborative filtering The source codes of CHESTNUT are online at https://githubcom/unnc-ucc/CHESTNUT
Citations
More filters
TL;DR: In this paper, a convolutional neural network (CNN) is integrated with the Particle Swarm Optimization (PSO) algorithm to generate serendipitous recommendations, which is based on the focus shift points, consisting of unexpectedness and relevance parameters.
Abstract: Most of the available recommender systems focus on the accuracy of recommendations. As a result, their recommendations are often popular and very close to user preferences, which make them repetitious and predictable, hence adversely affecting user satisfaction. Recent studies on recommender systems, however, aim for factors beyond accuracy as accuracy alone cannot ensure the satisfaction of all users. One of the most important criteria beyond the accuracy is serendipity, which includes relevant, unexpected, and novel recommendations that cannot be easily discovered by users themselves. In this paper, a Convolutional Neural Network (CNN) is integrated with the Particle Swarm Optimization (PSO) algorithm to generate serendipitous recommendations. The proposed method is based on the focus shift points, consisting of unexpectedness and relevance parameters. In this approach, these points are considered as the factors showing whether recommendations are serendipitous. The CNN is employed to predict the focus shift points for each user. Then, the PSO is utilized to search for recommendations close to the predicted focus shift points and generate the list of candidate recommendations. After that, the Serendipitous Personalized Ranking (SPR) method is employed to re-rank the candidate recommendations and generate the final list. According to the evaluation results, the proposed approach outperforms other state-of-the-art methods in SRDP, Hit Ratio, and NDCG factors.
6 citations
TL;DR: In this paper , the authors conducted an eight-day diary study with eight participants, using the YouTube recommender system, and found that users expect serendipitous recommender systems not only to provide surprising and diverse items, but also to guide the desired impact in the longterm perspectives and help them discover unrealized needs which do not rely on their past behaviours.
Abstract: AbstractPersonalized recommender systems have been criticized for limiting opportunities to consume diverse content and reinforcing self-bias leading to negative side effects on society. To address these issues, serendipity has emerged as a design goal of recommender systems increasing the long-term satisfaction of the users. However, research on serendipity in recommender systems has focused on improving the performance of algorithms to predict surprising and diverse items without considering the user’s desired experience. To investigate user expectations of serendipitous recommender systems, we conducted an eight-day diary study with eight participants, using the YouTube recommender system. We found that users expect serendipitous recommender systems not only to provide surprising and diverse items based on existing definitions of serendipity but also to guide the desired impact in the long-term perspectives and help them discover unrealized needs which do not rely on their past behaviours. Users also expect serendipitous recommender systems to provide items that could relieve their burden posed by the novelty of serendipitous recommendations. Based on these findings, we discuss design implications for designing user-centred serendipitous recommender systems that can support users to experience fundamental values of serendipity.KeywordsRecommender systemsSerendipityUser expectationsUser-centred design
19 Jul 2020
TL;DR: The preliminary feedback has shown that, compared with mainstream collaborative filtering techniques, though CHESTNUT limited users’ feelings of unexpectedness to some extent, it showed significant improvement in their feelings about certain metrics being both beneficial and interesting, which substantially increased users' experience of serendipity.
Abstract: Towards a serendipitous recommender system with user-centred understanding, we have built CHESTNUT, an Information Theory-based Movie Recommender System, which introduced a more comprehensive understanding of the concept. Although off-line evaluations have already demonstrated that CHESTNUT has greatly improved serendipity performance, feedback on CHESTNUT from real-world users through online services are still unclear now. In order to evaluate how serendipitous results could be delivered by CHESTNUT, we consequently designed, organized and conducted large-scale user study, which involved 104 participants from 10 campuses in 3 countries. Our preliminary feedback has shown that, compared with mainstream collaborative filtering techniques, though CHESTNUT limited users’ feelings of unexpectedness to some extent, it showed significant improvement in their feelings about certain metrics being both beneficial and interesting, which substantially increased users’ experience of serendipity. Based on them, we have summarized three key takeaways, which would be beneficial for further designs and engineering of serendipitous recommender systems, from our perspective. All details of our large-scale user study could be found at https://github.com/unnc-idl-ucc/Early-Lessons-From-CHESTNUT.
References
More filters
01 Apr 2001
TL;DR: This paper analyzes item-based collaborative ltering techniques and suggests that item- based algorithms provide dramatically better performance than user-based algorithms, while at the same time providing better quality than the best available userbased algorithms.
Abstract: Recommender systems apply knowledge discovery techniques to the problem of making personalized recommendations for information, products or services during a live interaction. These systems, especially the k-nearest neighbor collaborative ltering based ones, are achieving widespread success on the Web. The tremendous growth in the amount of available information and the number of visitors to Web sites in recent years poses some key challenges for recommender systems. These are: producing high quality recommendations, performing many recommendations per second for millions of users and items and achieving high coverage in the face of data sparsity. In traditional collaborative ltering systems the amount of work increases with the number of participants in the system. New recommender system technologies are needed that can quickly produce high quality recommendations, even for very large-scale problems. To address these issues we have explored item-based collaborative ltering techniques. Item-based techniques rst analyze the user-item matrix to identify relationships between di erent items, and then use these relationships to indirectly compute recommendations for users. In this paper we analyze di erent item-based recommendation generation algorithms. We look into di erent techniques for computing item-item similarities (e.g., item-item correlation vs. cosine similarities between item vectors) and di erent techniques for obtaining recommendations from them (e.g., weighted sum vs. regression model). Finally, we experimentally evaluate our results and compare them to the basic k-nearest neighbor approach. Our experiments suggest that item-based algorithms provide dramatically better performance than user-based algorithms, while at the same time providing better quality than the best available userbased algorithms.
8,634 citations
"CHESTNUT: Improve Serendipity in Mo..." refers methods in this paper
...We have also draw inspiration from the implementation and optimization of memory-based collaborative filtering techniques to enhance the system performance [9, 8, 27, 7]....
[...]
01 Aug 1999
2,686 citations
21 Apr 2006
TL;DR: This paper proposes informal arguments that the recommender community should move beyond the conventional accuracy metrics and their associated experimental methodologies, and proposes new user-centric directions for evaluating recommender systems.
Abstract: Recommender systems have shown great potential to help users find interesting and relevant items from within a large information space. Most research up to this point has focused on improving the accuracy of recommender systems. We believe that not only has this narrow focus been misguided, but has even been detrimental to the field. The recommendations that are most accurate according to the standard metrics are sometimes not the recommendations that are most useful to users. In this paper, we propose informal arguments that the recommender community should move beyond the conventional accuracy metrics and their associated experimental methodologies. We propose new user-centric directions for evaluating recommender systems.
1,072 citations
"CHESTNUT: Improve Serendipity in Mo..." refers background in this paper
...Within the Recommender System field, serendipity has been understood as receiving an unexpected and fortuitous item recommendation [20]....
[...]
...However, as a user-centric concept, serendipity has been understood narrowly within the Recommender System field, and it has been defined in previous research as receiving an unexpected and fortuitous item recommendation [20]....
[...]
26 Sep 2010
TL;DR: It is argued that the new ways of measuring coverage and serendipity reflect the quality impression perceived by the user in a better way than previous metrics thus leading to enhanced user satisfaction.
Abstract: When we evaluate the quality of recommender systems (RS), most approaches only focus on the predictive accuracy of these systems. Recent works suggest that beyond accuracy there is a variety of other metrics that should be considered when evaluating a RS. In this paper we focus on two crucial metrics in RS evaluation: coverage and serendipity. Based on a literature review, we first discuss both measurement methods as well as the trade-off between good coverage and serendipity. We then analyze the role of coverage and serendipity as indicators of recommendation quality, present novel ways of how they can be measured and discuss how to interpret the obtained measurements. Overall, we argue that our new ways of measuring these concepts reflect the quality impression perceived by the user in a better way than previous metrics thus leading to enhanced user satisfaction.
597 citations
08 Feb 2012
TL;DR: This work introduces the Auralist recommendation framework, a system that - in contrast to previous work - attempts to balance and improve all four factors simultaneously while limiting the impact on accuracy.
Abstract: Recommendation systems exist to help users discover content in a large body of items. An ideal recommendation system should mimic the actions of a trusted friend or expert, producing a personalised collection of recommendations that balance between the desired goals of accuracy, diversity, novelty and serendipity. We introduce the Auralist recommendation framework, a system that - in contrast to previous work - attempts to balance and improve all four factors simultaneously. 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 user satisfaction.
331 citations