PhilHuman's Objectives
Overview of the research
- Semantic web
- Natural language understanding
- Sentiment & emotion detection
- Multilinguality support
- Natural language generation
- Insights mining
- Creation of summaries and specific content
- Computer Vision & Machine Learning
- Context / Action / Object Recognition & Anticipation
- Facial Analysis for emotions understanding
- Body Pose Analysis
- First Person (Egocentric) Vision
- Conversational interfaces for home healthcare, chat bots, Q&A
- Integration of clinical knowledge, guidelines, data analytics, and counselling knowhow
- Application and exploitation of industrial use cases
- EU business plans development
- Economic technological aspects
8 Early Stage Researchers
16 Partners
8 Countries
Results
The PhilHumans project has trained a next generation of young researchers in innovative Artificial Intelligence (AI) and established user interaction with their personal health devices in an advanced and intuitive way. The project explored cutting-edge research topics related to AI-supported human-machine interfaces for personal health services. PhilHumans was committed to responsible research and innovation to establish disruptive and innovative technology for AI-assisted human-machines interfaces, employing language technology, cognitive computing, computer vision, and machine learning (ML). The technology can be applied in a number of personal health contexts and extend or being coupled with Home healthcare, as well as in additional fields such as population health management and provide several benefits to users making sure science and research is conducted with and for society.
The training and research network with 8 ESR in the project explored AI knowledge and expertise from Natural Language Generation (NLG) & Processing (NLP), Cognitive Computing, Computer Vision, ML focusing on 5 research objectives. Based on sound career development plans, and coached by experienced supervisors a training was offered by leading image analysis research groups from Philips (global leader in medical imaging) and the Eindhoven university of Technology (worldwide recognized authority in education and research on image analysis, esp. on MRI) and supported by researchers from leading universities like University of Cagliari, University of Catania and University of Aberdeen. After finalisation of their PhD the researchers planned for a next career step in research or industry depending on their affinity.
The project as a whole has contributed to the development of new technologies, data sets, and application concepts in the area of personal health interfaces. The work has been documented in the top conferences and journals in the area and many of the results, including data sets developed in the project, and software repositories, are available for the further research and development work in the community.
Consortium
Meet our partners
ESRs
Fellow’s individual research projects
Learn more
| ESR | Title Article | Authors | Year | Publisher | Type |
|---|---|---|---|---|---|
| ESR1 | Towards a Generalised Framework for Behaviour Insight Mining | Allmin Susaiyah, Aki Härmä, Ehud Reiter, Rim Helaoui and Milan Petković | 2020 | SmartPHIL: 1st Workshop on Smart Personal Health Interfaces | Conference/Workshop Paper |
| ESR | Iterative Neural Scoring of Validated Insight Candidates | Allmin Susaiyah, Aki Härmä, Ehud Reiter and Milan Petković | 2020 | Association for Computational Lingustics | Conference/Workshop Paper |
| ESR | Neural Scoring of Logical Inferences from Data using Feedback | Allmin Susaiyah, Aki Härmä, Ehud Reiter and Milan Petković | 2021 | Association for Computational Lingustics | Journal Paper |
NLP, semantics and sentiment analysis from text
YOUR TEAM: You will work in the Department of Mathematics and Computer Science of the University of Cagliari
GOAL: The overall goal of this thesis is to identify and understand the personal health of the user from the emotions expressed along the text or said from speech, for example in applications in Mother&Child care and healthy-living and assisted care. We aim at: (i) using ontology of personal health terms which includes terms related to body organs, symptoms, treatment, medical professional designations; (ii) using lexical and semantic resources to identify terms that hold only health-related meaning and more ambiguous terms; (iii) leveraging NLP services defined within RO1 and semantics to come up with a system for emotion detection and associated REST APIs. For such a research objective, NLP techniques and tools will be analysed to detect those that can be applied to the Sentiment Analysis problem in personal health. Supervised and unsupervised approaches will be identified to have a clear overview of the current state of art. Then, lexical resources and frameworks such as (not limited to) FrameNet, SentiWordNet, WordNet, VerbNet, FrameBase, BabelNet, FRED, FrameSter will be taken into account with the purpose of defining a multi-disciplinary approach to sentiment analysis at the cross-roads between affective computing and common sense computing focused in personal health. The overall goal is to better recognize, interpret and process opinions and sentiment out of a text document or speech of a certain user within the personal health domain.
ESR activities and expected results
- To bridge the semantic gap between word-level natural language data and the concept-level opinions conveyed by these so as to provide a more efficient passage from unstructured textual information to structured machine-processable data. For such a purpose, model ontologies (especially the existing ones in personal health domain) will be defined and developed using best practices of Semantic Web.
- Identify the opinions (statement by somebody, called the holder), the opinion holders, and facts/topics (e.g. ailment, treatment, medications, and other topics within the personal health domain) where the opinions are expressed on by representing the semantics of a sentence and modelling the roles played by its elements with respect to a model of opinion sentences.
- Defining a new Semantic Role Labelling exploiting new lexical and semantic resources in personal health context.
- Build a new emotion detection system focused in personal health on top of SENTILO, a Sentiment Analysis system that we recently built, that performs sentence-based sentiment analysis and relies on FRED, a machine reader for the Semantic Web. FRED relies on Combinatory Categorical Grammar, Discourse Representation Theory, Frame Semantics (e.g. Semantic Role Labelling), and Ontology Design Patterns and performs Named Entity Resolution, Coreference Resolution, Word Sense Disambiguation.
- Creation of new lexical resources for Sentiment Analysis (e.g. annotated dataset, word embeddings within the health domain, etc.)
- Provision of REST services for the provided Emotion detection system.
The fellow will build on several interconnected competences including automata theory, machine learning, statistics, probability models, and pattern recognition.
- Develop computable models of goal-oriented interactive health self-management systems;
- Develop hands-on experience with ML, data analysis, and optimization techniques for personal health applications and services containing sensor data, natural language content, and user interaction data;
- Acquire an in-depth understanding of the state-of-the-art ML techniques for Neural Abstract Machines (NAS) and program induction;
- Create novel computational methods that combine data-driven and knowledge-driven approaches and validating them in concrete personal health propositions;
- Explore, experiment and pilot project concepts creation for personal health
| Host | Timing | Length | Purpose |
|---|---|---|---|
| NUI Galway | M9 | 1-2 weeks | Training on big data in healthcare |
| UNICT | M12 | 1-2 weeks | Training on advanced machine learning algorithms |
| FBK | M25 | 1-2 weeks | Training on cognitive computing technology |
| R2M | M28 | 1-2 weeks | Training on Exploitation and Dissemination of Results |
