A5032082966- Complex Network Analysis Techniques
- Web Data Mining and Analysis
- Advanced Image and Video Retrieval Techniques
- Advanced Clustering Algorithms Research
- Text and Document Classification Technologies
- Advanced Text Analysis Techniques
- Image Retrieval and Classification Techniques
- Privacy, Security, and Data Protection
- Control Systems and Identification
- Balance, Gait, and Falls Prevention
- Spam and Phishing Detection
- Robot Manipulation and Learning
- EEG and Brain-Computer Interfaces
- Motor Control and Adaptation
- Network Security and Intrusion Detection
- Privacy-Preserving Technologies in Data
- Data Management and Algorithms
- Prosthetics and Rehabilitation Robotics
- Advanced Manufacturing and Logistics Optimization
- Video Analysis and Summarization
- Advanced Graph Neural Networks
- Robotic Mechanisms and Dynamics
- Neuroscience, Education and Cognitive Function
- Sentiment Analysis and Opinion Mining
- Robotics and Automated Systems
Information Technologies Institute
2012-2016
Centre for Research and Technology Hellas
2013-2016
Information Technology Institute
2014-2015
University of Edinburgh
2006-2007
Online social and news media generate rich timely information about real-world events of all kinds. However, the huge amount data available, along with breadth user base, requires a substantial effort filtering to successfully drill down relevant topics events. Trending topic detection is therefore fundamental building block monitor summarize originating from sources. There are wide variety methods variables they greatly affect quality results. We compare six on three Twitter datasets...
A large variety of features can be extracted from raw multimedia items. Moreover, in many contexts, like the case uploaded by users social media platforms, items may linked to metadata that very useful for a analysis tasks. Nevertheless, such are typically heterogeneous and difficult combine unified representation would suitable analysis. In this paper, we discuss problem clustering collections with purpose detecting events. order achieve this, novel multimodal algorithm is proposed. The...
Textual topic detection methods that work by clustering terms according to their cooccurrence patterns are called feature-pivot methods. Typically, the similarity measure is used for such takes into account of only pairs items. In this work, we argue examining simultaneous a larger number terms, better option when corpus contains set closely related fine-grained topics. To end, treat problem as Frequent Pattern Mining and propose novel algorithm "soft" Mining. We test proposed approach using...
The paper describes a multimodal graph-based system for addressing the Yahoo-Flickr Event Summarization Challenge of ACM Multimedia 2015. objective is to automatically uncover structure within collection 100 million photos/videos in form detecting and identifying events, summarizing them succinctly consumer consumption. presented uses sliding window over stream multimedia items build maintain same-event image graph applies clustering algorithm detect events. In addition, it makes use...
The phenomenal increase in the use of social media recent years has raised a number issues related to privacy. In this paper, we propose framework for raising awareness Online Social Network (OSN) users with respect information about them that is disclosed and can be inferred by OSN service operators as well third parties access their data. This takes form semantic, hierarchical scoring structure, enables easily browse over different privacy-related aspects presence network. Contrary...
Classic adaptive control methods for handling varying loads rely on an analytically derived model of the robot's dynamics. However, in many situations, it is not feasible or easy to obtain accurate analytic An alternative deriving dynamics learning from movement data. This paper describes a load estimation technique that uses learned instead We study examples where various inertial parameters are estimated models, their effectiveness evaluated along with robustness light imperfect,...
Adaptive motor control under continuously varying context, like the inertia parameters of a ma- nipulated object, is an active research area that lacks satisfactory solution. Here, we present and compare three novel strategies for learning context show how adding tactile sensors may ease this task. The first strategy uses only dynamics information to infer unknown parameters. It based on probabilistic generative model torques, which are linear in We demonstrate inference special case single...
Recent advances in machine learning and adaptive motor control have enabled efficient techniques for online of stationary plant dynamics it's use robust predictive control. However, realistic domains, system often change based on unobserved external contexts such as work load or contact conditions with other objects. Previous multiple model approaches to solving this problem are restricted finite, discrete without any generalization been tested only linear systems. We present a framework...
Online Social Networks (OSN) allow their users to organize friends into groups, also known as social circles. These circles can be used better manage who has access users' posted content and control the from other that they view. Unfortunately, these are generated manually this a laborious process for with more than few friends. In paper, we propose an approach automatically generating takes account both profile information of grouped network connectivity between them, while it allows...
This paper presents a real-time system that incorporates emerging knowledge from social media using crawling and mining techniques, developed within the SocialSensor FP7 project, are designed for addressing particularities of Web. The surfaced information is presented an interface optimized adapted to context user, taking into account efficient content delivery techniques optimize quality experience enhance user interaction. Research approaches discussed, particularly focusing in analytics,...
This paper describes a multimodal graph-based approach to address the problem of event detection and summarization in large scale image collections. A first version our system was presented Yahoo-Flickr Event Summarization Challenge ACM Multimedia 2015 [6]. The objective is automatically detect events within millions photos summarizing them efficiently for user consumption. uses moving time window over collection multimedia items build same-event graph applies clustering events. In addition,...