A5076503597- Biomedical Text Mining and Ontologies
- Bioinformatics and Genomic Networks
- Computational Drug Discovery Methods
- Service-Oriented Architecture and Web Services
- Caching and Content Delivery
- IoT and Edge/Fog Computing
- Machine Learning in Healthcare
- Sentiment Analysis and Opinion Mining
- Cloud Computing and Resource Management
- Recommender Systems and Techniques
- Peer-to-Peer Network Technologies
- Software System Performance and Reliability
- Multimedia Communication and Technology
- Complex Network Analysis Techniques
- Dementia and Cognitive Impairment Research
- Human Mobility and Location-Based Analysis
- Semantic Web and Ontologies
- Traffic Prediction and Management Techniques
- Advanced Text Analysis Techniques
- Topic Modeling
- Genetic Associations and Epidemiology
- Genomics and Rare Diseases
- Web Data Mining and Analysis
- E-Government and Public Services
- Business Process Modeling and Analysis
National Centre of Scientific Research "Demokritos"
2019-2025
Institute of Informatics of the Slovak Academy of Sciences
2021-2025
The American College of Greece
2024
Institute of Informatics & Telecommunications
2019-2023
National Technical University of Athens
2009-2019
Institute of Communication and Computer Systems
2012
Microblog content poses serious challenges to the applicability of traditional sentiment analysis and classification methods, due its inherent characteristics. To tackle them, we introduce a method that relies on two orthogonal, but complementary sources evidence: content-based features captured by n-gram graphs context-based ones polarity ratio. Both are language-neutral noise-tolerant, guaranteeing high effectiveness robustness in settings considering. ensure our approach can be integrated...
Sentiment Analysis over Social Media facilitates the extraction of useful conclusions about average public opinion on a variety topics, but poses serious technical challenges. This is because sparse, noisy, multilingual content that posted on-line by users. In this paper, we introduce novel method for capturing textual patterns inherently supports challenging type content. essence, it creates graph whose nodes correspond to character n-grams document, while its weighted edges denote distance...
Abstract Background Dementia develops as cognitive abilities deteriorate, and early detection is critical for effective preventive interventions. However, mainstream diagnostic tests screening tools, such CAMCOG MMSE, often fail to detect dementia accurately. Various graph-based or feature-dependent prediction progression models have been proposed. Whenever these exploit information in the patients’ Electronic Medical Records, they represent promising options identify presence severity of...
Tailoring personalized treatments demands the analysis of a patient’s characteristics, which may be scattered over wide variety sources. These features include family history, life habits, comorbidities, and potential treatment side effects. Moreover, services visited most by patient before new diagnosis, as well type requested tests, uncover patterns that contribute to earlier disease detection effectiveness. Built on knowledge-driven ecosystems, we devise DE4LungCancer, health data...
In this paper we present a SOA (Service Oriented Architecture)-based platform, enabling the retrieval and analysis of big datasets stemming from social networking (SN) sites and Internet Things (IoT) devices, collected by smart city applications socially-aware data aggregation services. A large set city applications in areas Participating Urbanism, Augmented Reality Sound-Mapping throughout participating cities is being applied, resulting into produced sets millions...
This paper applies different link prediction methods on a knowledge graph generated from biomedical literature, with the aim to compare their ability identify unknown drug-gene interactions and explain predictions. Identifying novel drug-target is crucial step in drug discovery repurposing. One approach this problem predict missing links between gene nodes, that contains relevant knowledge. Such can be extracted using text mining tools. In work, we state-of-the-art embedding approaches...
Tackling the challenges posed by Social Networking content and addressing its casual nature, n-gram graphs technique provides a language-independent supervised approach for text mining. Adopting this data analysis model, paper an extended study of sentiment analysis, using multilingual multi-topic environment, employing combining different classification algorithms, attempting various configuration approaches on parameters to increase efficiency. Compared results found big corpora used in...
Microblog content poses serious challenges to the applicability of sentiment analysis, due its inherent characteristics. We introduce a novel method relying on content-based and context-based features, guaranteeing high effectiveness robustness in settings we are considering. The evaluation our methods over large Twitter data set indicates significant improvements traditional techniques.
The vision of IASIS project is to turn the wave big biomedical data heading our way into actionable knowledge for decision makers. This achieved by integrating from disparate sources, including genomics, electronic health records and bibliography, applying advanced analytics methods discover useful patterns. goal large amounts available information authorities planning public activities policies. integration analysis these heterogeneous sources will enable best decisions be made, allowing...
Mini-Mental State Examination (MMSE) is used as a diagnostic test for dementia to screen patient’s cognitive assessment and disease severity. However, these examinations are often inaccurate unreliable either due human error or patients’ physical disability correctly interpret the questions well motor deficit. Erroneous data may lead wrong of specific patient. Therefore, other clinical factors (e.g., gender comorbidities) existing in electronic health records, can also play significant role,...
In this paper we present the RADICAL platform, a software stack that enables combination of social network (SN) services and Internet Things (IoT) in context innovative smart cities. makes possible development deployment interoperable pervasive multi-sensory socially-aware services; facilitates governance flexible replication across cities regions through Virtual Machine generation mechan-ism sophisticated cloud environment. A large scale piloting platform integrates, deploys tests various...
The Social Networking Sites (SNS) comprise a pool from which developers can pump functionality and data. Usable REST APIs are providing access to two valuable business assets: the users' generated content social graph. lack of standards antagonistic nature SNSs have resulted in use proprietary API specifications -in turndata models. Each SNS uses different method way describe notions largely similar, e.g. “friends” or shared “multimedia items”. conceptual similarity between entities “living”...
Early and precise prognosis of dementia is a critical medical challenge. The design an optimal computational model that addresses this issue, at the same time explains underlying mechanisms lead to output decisions, ongoing In study, we focus on assessing risk individual converting Dementia in short (next year) long (one five years) term, given only few early-stage observations. Our goal develop machine learning could assist prediction from regular clinical data. results show combining...
With the emergence of service provisioning environments and new networking capabilities, antagonistic businesses have been able to collaborate securely by sharing information in order a beneficial result for all. This collaboration has sometimes imposed state legislation desirable firms themselves so as resolve frequently occurring abnormalities. In any case, exchange takes place between firms, security privacy issues arise. context this paper, collaborative environment analyzed enterprises...
Edge computing has seen a great progress nowadays eliminating the network latency risks by placing data and functionality in low-end devices closer to end user is central application domains such as support of backend mobile apps. When this problem combined with big requirements linked number app users, optimization edge resource utilization becomes extremely important cost- quality-wise. This work explores approaches model demand based on mobility characteristics. The Application User...
Deciding and reserving appropriate resources in the Cloud, is a basic initial step for adopters when employing an Infrastructure as Service to host their application. However, size number of Virtual Machines used, along with expected application workload, will highly influence its operation, terms observed Quality Service. This paper proposes machine learning approach, based on Artificial Neural Networks, mapping required levels (expected) workload concrete resource demands. The presented...