A5079385763- Recommender Systems and Techniques
- Advanced Bandit Algorithms Research
- Machine Learning and Data Classification
- Advanced Text Analysis Techniques
- Topic Modeling
- Opinion Dynamics and Social Influence
- Business Process Modeling and Analysis
- Machine Learning in Healthcare
- Social and Intergroup Psychology
- Ethics and Social Impacts of AI
- Social Media and Politics
- Complex Network Analysis Techniques
- Speech and dialogue systems
- Data Stream Mining Techniques
- Customer churn and segmentation
- Access Control and Trust
- Personal Information Management and User Behavior
- Privacy, Security, and Data Protection
- Privacy-Preserving Technologies in Data
- Cultural Differences and Values
- Decision-Making and Behavioral Economics
- Psychological Well-being and Life Satisfaction
- Information Retrieval and Search Behavior
- Sentiment Analysis and Opinion Mining
- Explainable Artificial Intelligence (XAI)
University of Antwerp
2022-2024
Vrije Universiteit Brussel
2024
Victoria University of Bangladesh
2010
This report documents the program and outcomes of Dagstuhl Seminar 23031 "Frontiers Information Access Experimentation for Research Education", which brought together 38 participants from 12 countries. The seminar addressed technology-enhanced information access (information retrieval, recommender systems, natural language processing) specifically focused on developing more responsible experimental practices leading to valid results, both research as well scientific education. featured a...
News media play an important role in democratic societies. Central to fulfilling this is the premise that users should be exposed diverse news. However, news recommender systems are gaining popularity on websites, which has sparked concerns over filter bubbles. More specifically, editors, policy-makers and scholars worried these may expose less content time. To best of our knowledge, hypothesis not been tested a longitudinal observational study real interact with website. Such studies...
RecPack is an easy-to-use, flexible and extensible toolkit for top-N recommendation with implicit feedback data. Its goal to support researchers the development of their algorithms, from similarity-based deep learning allow correct, reproducible reusable experimentation. In this demo, we give overview package show how can use it advantage when developing algorithms.
Information access systems, such as Google News or YouTube, increasingly employ algorithms to rank diverse content music, recipes, and news articles. Acknowledging the influential role of these gatekeepers online content, research community is exploring 'beyond-accuracy' metrics. However, deciding what norms values are relevant should be prioritized when designing evaluating information systems a challenging task. This tutorial aims cultivate normative thinking decision-making in design...
Recommender systems are among the most widely used applications of artificial intelligence. Since they so used, it is important that we, as practitioners and researchers, think about impact these may have on users, society, other stakeholders. To effect, NORMalize workshop seeks to introduce normative thinking, consider norms values underpin recommender in community. The objective bring together a growing community researchers across disciplines who want should be considered design...
Previous research has used Large Language Models (LLMs) to develop personalized Conversational Recommender Systems (CRS) with text-based user interfaces (UIs). However, the potential of LLMs generate interactive graphical elements that enhance experience remains largely unexplored. To address this gap, we introduce "GenUI(ne) CRS," a novel framework designed leverage for adaptive and UIs. Our supports domain-specific such as buttons cards, in addition inputs. It also addresses common LLM...
Evaluating recommender systems adequately and thoroughly is an important task. Significant efforts are dedicated to proposing metrics, methods, protocols for doing so. However, there has been little discussion in the systems’ literature on topic of testing. In this work, we adopt adapt concepts from software testing domain, e.g., code coverage, metamorphic testing, or property-based help researchers detect correct faults recommendation algorithms. We propose a test suite that can be used...
When designing recommender-systems experiments, a key question that has been largely overlooked is the choice of datasets. In brief survey ACM RecSys papers, we found authors typically justified their dataset choices by labelling them as public, benchmark, or 'real-world' without further explanation. We propose Algorithm Performance Space (APS) novel method for informed selection. The APS an n-dimensional space where each dimension represents performance different algorithm. Each depicted...
Recommender systems are among the most widely used applications of artificial intelligence. Their use can have far-reaching consequences for users, stakeholders, and society at large. In this second edition NORMalize workshop, we once again seek to advance research agenda normative thinking, considering norms values that underpin recommender systems, as well introduce concept a broader audience. We aim bring together growing community researchers practitioners across disciplines who want...
With the increasing use of AI and ML-based systems, interpretability is becoming an increasingly important issue to ensure user trust safety. This also applies area recommender where methods based on matrix factorization (MF) are among most popular for collaborative filtering tasks with implicit feedback. Despite their simplicity, latent factors users items lack in case effective, unconstrained MF-based methods. In this work, we propose extended Dirichlet Allocation model (LDAext) that has...