The 2nd Workshop on Intelligent Recommender Systems by Knowledge Transfer & Learning

recsysktl 2018


Data Mining & Analysis



Call For Papers
Home
Recommender systems provide relevant items and information to users by profiling users and items in various ways. Growth of online information systems has led to an abundance of data that is heterogeneous, noisy, and changes rapidly. The data used by recommender systems, in forms of implicit or explicit user feedback, follow the same trend: the feedback can be in various formats, such as ratings, online behaviors, or textual reviews, and collected from multiple resources, the collected feedback is uncertain, and user taste and item popularities can change over time. In this workshop, the focus is on recommender systems’ data heterogeneity: collected feedback with various types, collected from various domains, contexts, or applications.
While the data heterogeneity provides multiple views to users’ preferences, and thus, may be helpful in recommending more related items to users, it may also add more noise and uncertainty to the data and lead to weaker recommendations. Cross-domain recommender systems and transfer learning approaches propose to effectively take advantage of such diversity of viewpoints to provide better-quality recommendations and resolve issues such as the cold-start problem.
The emerging research on cross-domain, context-aware and multi-criteria recommender systems, has proved to be successful. Given the recent availability of cross-domain datasets and novelty of the topic, we organize the 2nd workshop on intelligent recommender systems by knowledge transfer and learning (RecSysKTL) held in conjunction with ACM RecSys 2018. This workshop intends to create a medium to generate more practical and efficient predictive models or recommendation approaches by leveraging user feedbacks or preferences from multiple domains.
---------------------------------------------------------------------------------------------------
Generally, we focus on the topic of “cross-domain”, where the notion of “domain” may vary from applications to applications. For example, the concept of context-aware and multi-criteria recommender systems can also be considered as an application of “cross-domain” techniques. Particularly, we are interested in how to apply knowledge transfer and learning approaches to build intelligent recommender systems. Domain could be (but not limited to):
From one application to another: We may utilize user behaviors on social networks to predict their preferences on movies (e.g., Netflix, Youtube) or music (e.g., Pandora, Spotify).
From one category to another: We may predict a user’s taste on electronics by using his or her preference history on books based on the data collected from Amazon.com.
From one context to another: We may collect a user’s preferences on the items over different time segment (e.g., weekend or weekday) and predict her preferences on movie watching within another context (e.g., companion and location).
From one task to another: It may be useful for us to predict how a user will select hotels for his or her vocations by learning from how he or she books the tickets for transportations.
From one structure to another: It could be also possible for us to infer social connections by learning from the structure of heterogeneous information networks.
The topics of interest for this workshop include (but are not limited to):
Applications of Knowledge Transfer for Recommender Systems
Cross-domain recommendation
Context-aware recommendation, time-aware recommendation
Multi-criteria recommender systems
Novel applications
Methods for Knowledge Transfer in Recommender Systems
Knowledge transfer for content-based filtering
Knowledge transfer in user- and item-based collaborative filtering
Transfer learning of model-based approaches to collaborative filtering
Deep Learning methods for knowledge transfer
Challenges in Knowledge Transfer for Recommendation
Addressing user feedback heterogeneity from multiple domains (e.g. implicit vs. explicit, binary vs. ratings, etc.)
Multi-domain and multi-task knowledge representation and learning
Detecting and avoiding negative (non-useful) knowledge transfer
Ranking and selection of auxiliary sources of knowledge to transfer from
performance and scalability of knowledge transfer approaches for recommendation
Evaluation of Recommender Systems based on Knowledge Transfer
Beyond accuracy: novelty, diversity, and serendipity of recommendations supported by the transfer of knowledge
Performance of knowledge transfer systems in cold-start scenarios
Impact of the size and quality of transferred data on target recommendations
Analysis of the amount of domain overlap on recommendation performance
Submissions Guidelines
We accept long papers (up to 8 pages) and short papers (up to 4 pages) in ACM conference format (references are counted in the page limit). Long papers are expected to present original research work which should report on substantial contributions of lasting value. Short papers may discuss the late-breaking results or exciting new work that is not yet mature, or open challenges in promising research directions. The accepted papers will be invited for presentations and the proceedings will be available at http://ceur-ws.org, while the authors will hold the copyrights.
Note: the review process is double-blind. All submissions must be appropriately anonymized according to the following guidelines:
Author’s names and affiliations are not visible anywhere in the paper.
Acknowledgements should be anonymized or removed during the review process.
Self-citations should be included where necessary, but must use the third person. For example, “… as shown in our previous user study [2] … ” is not allowed, whereas “… as shown in Smith et al. [2] ” is acceptable (because in this case the citation [2] will NOT be perceived as self-citation).
All of the submissions should be submitted via EasyChair system: https://easychair.org/conferences/?conf=recsysktl2018
Please note that authors cannot be added after acceptance of papers and at least one of the paper authors is expected to attend the workshop in person to present their papers.
We are working on a special issue, and the authors will be invited to submit the extension of their work to the special issue in a journal. More information will be released later.