A5091680688- Topic Modeling
- Advanced Text Analysis Techniques
- Semantic Web and Ontologies
- Sentiment Analysis and Opinion Mining
- Natural Language Processing Techniques
- Web Data Mining and Analysis
- Human Mobility and Location-Based Analysis
- Urban Transport and Accessibility
- Heart Rate Variability and Autonomic Control
- Recommender Systems and Techniques
- Information Retrieval and Search Behavior
- Transportation Planning and Optimization
- Music and Audio Processing
- Biomedical Text Mining and Ontologies
- Academic integrity and plagiarism
- Non-Invasive Vital Sign Monitoring
- Spam and Phishing Detection
- Emotion and Mood Recognition
- Service-Oriented Architecture and Web Services
- AI in Service Interactions
- Mobile Crowdsensing and Crowdsourcing
- Data Mining Algorithms and Applications
- Speech and Audio Processing
- Imbalanced Data Classification Techniques
- Advanced Wireless Network Optimization
Universitat Pompeu Fabra
2013-2024
Barcelona Media
2008-2012
Laboratoire Informatique d'Avignon
2004-2006
In this paper we present a study on music mood classification using audio and lyrics information. The of song is expressed by means musical features but relevant part also seems to be conveyed the lyrics. We evaluate each factor independently explore possibility combine both, natural language processing information retrieval techniques. show that standard distance-based methods latent semantic analysis are able classify significantly better than random, performance still quite inferior...
While collaborative filtering often yields very good recommendation results, in many real-world scenarios cold-start and data sparseness remain important problems. This paper presents a hybrid recommender system that integrates user demographics item characteristics, around core based on user-item interactions. The is evaluated Movie lens (including genre information data) as well from discount coupon provider. We show the inclusion of additional can have great impact quality, especially...
Stress can be considered a mental/physiological reaction in conditions of high discomfort and challenging situations. The levels stress reflected both the physiological responses speech signals person. Therefore study fusion two modalities is great interest. For this cause, public datasets are necessary so that different proposed solutions comparable. In work, publicly available multimodal dataset for detection introduced, including cues data. include electrocardiograph (ECG), respiration...
Flood management and media production planning are both tasks that require timely sound decision making, as well effective collaboration between professionals in a team split remote headquarter operators situ actors. This paper presents an extended reality (XR) platform utilizes interactive immersive technologies integrates artificial intelligence (AI) algorithms to support the public involved such incidents events. The developed XR tools address various specialized end-user needs of...
Abstract Stress can be considered a mental/physiological reaction in conditions of high discomfort and challenging situations. The levels stress reflected both the physiological responses speech signals person. In this work, we introduce novel decision-level fusion framework for multimodal level detection based on from wearable devices user audio recordings. include Electrocardiograph (ECG), Respiration (RSP), InertialMeasurement Unit (IMU) sensors equipped smart vest. A data collection...
This paper provides an analysis of human mobility data in urban area using the amount available bikes stations community bicycle program Bicing Barcelona. The was obtained by periodic mining a KML-file accessible through website. Although principle very noisy, after some preprocessing and filtering steps allows to detect temporal patterns as well identify residential, university, business leisure areas city. results lead proposal for improvement bicing website, including prediction number...
Apart from being an economic struggle, migration is first of all a societal challenge; most migrants come different cultural and social contexts, do not speak the language host country, are familiar with its societal, administrative, labour market infrastructure. This leaves them in need dedicated personal assistance during their reception integration. However, due to continuously high number people attendance, public administrations non-governmental organizations often overstrained by this...
Jens Grivolla, Maite Melero, Toni Badia, Cosmin Cabulea, Yannick Estève, Eelco Herder, Jean-Marc Odobez, Susanne Preuß, Raúl Marín. Proceedings of the Workshop on Open Infrastructures and Analysis Frameworks for HLT. 2014.