A5039366726- Recommender Systems and Techniques
- Advanced Bandit Algorithms Research
- Video Analysis and Summarization
- Image Retrieval and Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Music and Audio Processing
- Consumer Market Behavior and Pricing
- Topic Modeling
- IoT and Edge/Fog Computing
- Multimedia Communication and Technology
- Digital Marketing and Social Media
- Machine Learning and Algorithms
- Visual Attention and Saliency Detection
- Intelligent Tutoring Systems and Adaptive Learning
- Big Data and Business Intelligence
- Data Visualization and Analytics
- Advanced Vision and Imaging
- Sentiment Analysis and Opinion Mining
- Mobile Crowdsensing and Crowdsourcing
- Data Stream Mining Techniques
- Social Media and Politics
- Innovative Human-Technology Interaction
- Data Management and Algorithms
- Emotion and Mood Recognition
- Big Data Technologies and Applications
University of Bergen
2020-2024
Norwegian School of Economics
2023
Politecnico di Milano
2015-2022
Nottingham Trent University
2022
Free University of Bozen-Bolzano
2011-2020
Islamic Azad University, Mashhad
2016
Instituto Politécnico Nacional
2015
Stockholm University
2009-2010
Abstract Recommender systems help people find relevant content in a personalized way. One main promise of such is that they are able to increase the visibility items long tail , i.e., lesser-known catalogue. Existing research, however, suggests many situations today’s recommendation algorithms instead exhibit popularity bias meaning often focus on rather popular their recommendations. Such may not only lead limited value recommendations for consumers and providers short run, but it also...
Music recommender systems (MRSs) have experienced a boom in recent years, thanks to the emergence and success of online streaming services, which nowadays make available almost all music world at user’s fingertip. While today’s MRSs considerably help users find interesting these huge catalogs, MRS research is still facing substantial challenges. In particular when it comes build, incorporate, evaluate recommendation strategies that integrate information beyond simple user–item interactions...
Due to the extensive growth of food varieties, making better and healthier choices becomes more complex. Most current suggestion applications offer just generic advices that are not tailored user's personal taste. To tackle this issue, we propose in paper a novel recommender system provides high quality personalized recipe suggestions. These recommendations generated by leveraging data set users' preferences expressed form ratings tags, which signal food's ingredients or features users like....
Abstract The last two decades have witnessed major disruptions to the traditional media industry as a result of technological breakthroughs. New opportunities and challenges continue arise, most recently rapid advance adoption artificial intelligence technologies. On one hand, broad these technologies may introduce new for diversifying offerings, fighting disinformation, advancing data-driven journalism. other techniques such algorithmic content selection user personalization can risks...
Hybrid recommender systems utilize advanced algorithms capable of learning heterogeneous sources data and generating personalized recommendations for users. The can range from user preferences (e.g., ratings or reviews) to item content description category). Prior studies in the field have primarily relied on "ratings" as feedback, when building profiles evaluating quality recommendation. While are informative, they may still fail represent a comprehensive picture actual preferences. In...
The accuracy of collaborative-filtering recommender systems largely depends on three factors: the quality rating prediction algorithm, and quantity available ratings. While research in field often concentrates improving algorithms, even best algorithms will fail if they are fed poor-quality data during training, that is, garbage in, out. Active learning aims to remedy this problem by focusing obtaining better-quality more aptly reflects a user's preferences. However, traditional evaluation...
The ACM Recommender Systems Challenge 20171 focused on the problem of job recommendations: given a new advertisement, goal was to identify those users who are both (a) interested in getting notified about and (b) appropriate candidates for job. Participating teams had balance between user interests requirements as well dealing with cold-start situation. For first time history conference, RecSys challenge offered an online evaluation: compete part traditional offline evaluation top 25 were...
Abstract Reading or viewing recommendations are a common feature on modern media sites. What is shown to consumers as nowadays often automatically determined by AI algorithms, typically with the goal of helping discover relevant content more easily. However, highlighting filtering information that comes such may lead undesired effects even society, for example, when an algorithm leads creation filter bubbles amplifies spread misinformation. These well-documented phenomena create need...
In the last years, there has been much attention given to semantic gap problem in multimedia retrieval systems. Much effort devoted bridge this by building tools for extraction of high-level, semantics-based features from content, as low-level are not considered useful because they deal primarily with representing perceived content rather than semantics it.
In the task of modeling user preferences for movie recommender systems, recent research has demonstrated benefits describing movies with their eudaimonic and hedonic scores (E H scores), which reflect depth message level fun experience they provide, respectively. So far, labeling E been done manually using a dedicated instrument (a questionnaire), is time-consuming. To address this issue, we propose an automatic approach predicting scores. Specifically, collected 709 from 370 users (with...