About this Research Topic
The behavior of users in the digital world (e.g. online shopping, social media activity, etc.) is increasingly supported by recommender systems. Recommender systems are mainly data-driven, based on behavioral data, such as ratings, likes, purchases or general interaction and consumption. Although these systems are useful for both users and service providers, they have several drawbacks including the cold start problem (i.e., the data sparsity in the initial stages of system deployment), various biases resulting from biases in the user-generated data (i.e., gender, popularity, or selection bias) or the limited explainability of the data (i.e., using data without understanding the root cause of behaviors). Hence, recent work has started to adopt approaches that include sophisticated user analysis and modeling as well as algorithms that reduce biases and generate fair and explainable recommendations. Frequently, these systems take advantage of psychological models to explain and predict user interactions with the systems and allow for a deeper understanding of user behavior, preferences, and needs, which in turn also allow for more generalizable results. In complement, digital behavior in other systems has also been used to infer user characteristics. For example, social media activities have been used to analyze, model, and predict user behavior in recommender systems.
This Research Topic aims at bringing together state-of-the-art research that focuses on empirical analysis of user interactions with systems to improve our understanding of individual and collective user behavior, explore novel user models and novel recommendation and personalization algorithms, or focus on the understanding of mutual interdependence between user behavior and algorithms. We encourage authors to submit original research articles, case studies, reviews, methodological, theoretical and critical perspectives, and viewpoint articles within the usage of user behavior analysis and user models in all kinds of recommender systems on topics including, but not limited to:
● User analysis and models that explain online behavior:
○ Personality
○ Emotions/Mood and user needs
○ Cognitive biases
○ User preference elicitation
○ User perception of recommender systems
● Methods for Analyzing User Behavior:
○ Data-driven approaches to user analysis
○ Qualitative Methods to Collect and Analyze Behavioral data
○ Observational studies and natural experiments
○ Causal inference
○ Novel data science methods for the analysis of user behavior
● Recommender Systems and Algorithms:
○ Recommendation algorithms
○ Recommender and personalization system evaluation
○ Music, video and media recommendation
○ Recommendation of learning materials
○ Recommendations on social media platforms
○ Social recommenders
○ Interfaces for recommender systems
○ Novel machine learning approaches to recommendation algorithms
○ Conversational recommender systems
● Algorithmic Fairness and Transparency:
○ Fairness-aware recommendations
○ Diversity for recommendation systems
○ Explainable methods for recommendations
○ Group-fairness in recommender systems
Keywords: User modeling and recommendations, recommender systems, Algorithmic Fairness, Recommendation algorithms, personalization system evaluation, Social recommenders, Conversational recommender systems, Fairness-aware recommendations
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.