Program

Emotions constitute an important aspect when listening to music. While quality annotations provide a very well-defined and fine-grained description of the emotional content of music tracks, user-generated tags provide an alternative view stemming from large-scale data. In this work, we examine the relationship between these two characterizations of the emotional content of music. We compare the emotional profile of 453 music tracks as measured with annotations collected using tools based on psychological research on music and emotion, with the emotional profile extracted from user-generated data collected on music streaming platforms. Our analysis shows that these profiles often differ, therefore providing complementary views on the emotional content of music that should be leveraged simultaneously in emotion-based music information retrieval.

In the world of E-Commerce, the core task is to understand the personalized preference from various kinds of heterogeneous information, such as textual reviews, item images and historical behaviors. In current systems, these heterogeneous information are mainly exploited to generate better item or user representations. For example, in scenario of visual search, the importance of modeling query image has been widely acknowledged. But, these existing solutions focus on improving the representation quality of the query image, overlooking the personalized visual preference of the user. Note that the visual features affect the user’s decision significantly, e.g., a user could be more likely to click the items with her preferred design. Hence, it is fruitful to exploit the visual preference to deliver better capacity for personalization. To this end, we propose a simple yet effective target-aware visual preference learning framework (named Tavern) for both item recommendation and search. The proposed Tavern works as an individual and generic model that can be smoothly plugged into different downstream systems. Specifically, for visual preference learning, we utilize the image of the target item to derive the visual preference signals for each historical clicked item. This procedure is modeled as a form of representation disentanglement, where the visual preference signals are extracted by taking off the noisy information irrelevant to visual preference from the shared visual information between the target and historical items. During this process, a novel selective orthogonality disentanglement is proposed to avoid the significant information loss. Then, a GRU network is utilized to aggregate these signals to form the final visual preference representation. Extensive experiments over three large-scale real-world datasets covering visual search, product search and recommendation well demonstrate the superiority of our proposed Tavern against existing technical alternatives. Further ablation study also confirms the validity of each design choice.

A well-designed document communicates not only through its words but also through its visual eloquence. Authors utilize aesthetic elements such as colors, fonts, graphics, and layouts to shape the perception of information. Thoughtful document design, informed by psychological insights, enhances both the visual appeal and the comprehension of the content. While state-of-the-art document AI models demonstrate the benefits of incorporating layout and image data, it remains unclear whether the nuances of document aesthetics are effectively captured. To bridge the gap between human cognition and AI interpretation of aesthetic elements, we formulated hypotheses concerning AI behavior in document understanding tasks, specifically anchored in document design principles. With a focus on legibility and layout quality, we tested four aspects of aesthetic effects: noise, contrast, alignment, and complexity, on model confidence using correlational analysis. The results and observations highlight the value of model analysis rooted in document design theories. Our work serves as a trailhead for further studies and we advocate for continued research in this area to deepen our understanding of how AI interprets document aesthetics.

In recent years, the field of recommendation systems has witnessed significant advancements, with explainable recommendation systems gaining prominence as a crucial area of research. These systems aim to enhance user experience by providing transparent and compelling recommendations, accompanied by explanations. However, a persistent challenge lies in addressing biases that can influence the recommendations and explanations offered by these systems. Such biases often stem from a tendency to favor popular items and generate explanations that highlight their common attributes, thereby deviating from the objective of delivering personalized recommendations and explanations. While existing debiasing methods have been applied in explainable recommendation systems, they often overlook the model-generated explanations in tackling biases. Consequently, biases in model-generated explanations may persist, potentially compromising system performance and user satisfaction. To address biases in both model-generated explanations and recommended items, we discern the impact of model-generated explanations in recommendation through a formulated causal graph. Inspired by this causal perspective, we propose a novel approach termed Causal Explainable Recommendation System (CERS), which incorporates model-generated explanations into the debiasing process and enacts causal interventions based on user feedback on the explanations. By utilizing model-generated explanations as intermediaries between user-item interactions and recommendation results, we adeptly mitigate the biases via targeted causal interventions. Experimental results demonstrate the efficacy of CERS in reducing popularity bias while simultaneously improving recommendation performance, leading to more personalized and tailored recommendations. Human evaluation further affirms that CERS generates explanations tailored to individual users, thereby enhancing the persuasiveness of the system.