We innovate today for the solutions of tomorrow
Dr. Despina Elisavet Filippidou
Head of Innovation Start Hub
Email Address
[email protected]
Phone
+30 2310 500 181
DOTSOFT’s Research and Innovation Department is the key driver of the company’s technological advancement and differentiation, shaping the future of the solutions it offers. Through continuous participation in European and international research programs, the company maintains an active presence in the innovation ecosystem and integrates the latest technological developments into its services.
The Department operates as a dynamic knowledge hub, where research, technology, and business application come together to create new products and services. Through collaborations with universities, research institutions, and international organizations, access to cutting‑edge expertise is ensured, enabling the development of high‑value solutions.
Our activities include:
Publications – Presentations – Case Studies
Knowledge Graphs and Machine Learning in Fake News and Misinformation Detection
- The rise of digital platforms has accelerated the spread of misinformation, creating significant social, political, and economic challenges. Knowledge Graphs (KGs) are emerging as effective tools for improving the accuracy, interpretability, and scalability of fake‑news detection systems, addressing limitations of traditional machine‑learning approaches that rely mainly on linguistic analysis.
This work includes a literature review synthesizing findings from recent studies on the application of Knowledge Graphs in misinformation detection. It examines how KGs enhance detection by capturing real‑world relationships, analyzing context, and improving model interpretability, while also discussing existing limitations related to scalability, data completeness, and adaptation to specific contexts.
The reviewed studies highlight the need for future research focused on developing scalable, real‑time, and cross‑lingual Knowledge Graph models to strengthen global misinformation‑detection capabilities.
Additionally, preliminary results from two use cases are presented, showcasing a methodology for constructing Knowledge Graphs that can serve as a valuable tool for addressing the spread of misinformation.
Dynamic Demand Forecasting with AI: Maximizing Value in the Supply Chain
- This work presents the development of a dynamic demand‑forecasting platform designed to accurately predict demand and quantify its added value in supply‑chain management, leveraging advanced machine‑learning algorithms and operational management techniques.
The platform collects and processes real‑time data from multiple sources, such as point‑of‑sale systems, e‑commerce transactions, competitor pricing policies, weather conditions, and consumer‑sentiment data. By integrating environmental and product‑related parameters, the platform optimizes inventory levels, delivery routing, and overall operational efficiency.
The solution achieves high forecasting accuracy by utilizing models such as XGBoost and Support Vector Regression (SVR), which consistently outperform traditional forecasting methods. At the same time, it is designed to be scalable, cost‑effective, and user‑friendly, making it particularly suitable for small and medium‑sized enterprises (SMEs), which often face technological and financial constraints when adopting advanced forecasting systems.
The implementation of the system significantly enhances decision‑making processes, leading to reduced operational costs and improved customer satisfaction.
Real‑World Case Study on Sales Forecasting Using Artificial Intelligence
- 28th Panhellenic Conference on Informatics (PCI), 13–15 December 2024, Athens, Greece
Accurate sales forecasting is a critical factor for optimizing operations and supporting strategic decision‑making in the e‑commerce sector. This work presents a comprehensive review of existing literature on sales‑forecasting algorithms, covering both traditional statistical approaches and advanced artificial intelligence (AI) techniques.
Using a dataset from one of the largest e‑commerce companies in Greece, multiple AI‑based models are applied to predict sales trends. The preliminary results highlight the effectiveness of AI algorithms in identifying complex patterns within the data, achieving higher forecasting accuracy compared to conventional methods—particularly when semantic data from external sources, such as weather conditions and Google Trends, are incorporated.
These findings underscore the potential of integrating artificial intelligence into sales‑forecasting processes, contributing to enhanced competitiveness and operational efficiency for e‑commerce businesses.
White Paper: Cancer Survivorship and AI for Well‑Being – A Collaborative Approach of European Research Projects
- A White Paper has been published presenting the results of the collaboration among 11 CORDIS projects, offering valuable recommendations and directions for future research in the fields of well‑being, cancer, and artificial intelligence in health.
This document is the outcome of synergies among multiple Horizon 2020 projects participating in the cluster “Cancer Survivorship and AI for Wellbeing (CS‑AIW)”, aiming to promote innovative approaches that enhance the quality of life and care of cancer survivors.
The BD4QoL project made a substantial contribution, particularly in Chapter 4, titled “Lessons learned from our collaboration”, which focuses on the insights and conclusions derived from the cooperation among the research projects.
Smartphone‑Based Strategy for Monitoring Quality of Life in Head and Neck Cancer Survivors
- Part of the book series Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering (LNICST, Volume 488)
- apply self‑management practices through active symptom tracking,
maintain healthy lifestyle habits,
interact with embedded artificial intelligence that serves as an additional communication channel between patient and clinician,
- complete standardized QoL questionnaires via a web platform from home.
The challenges encountered, the selected non‑invasive data‑collection procedures, and the qualitative data derived from physical, social, and behavioral indicators together form a valuable set of guidelines and requirements for future research efforts focused on post‑treatment cancer‑patient monitoring using portable IoT devices.
Smartphone‑based IoT systems have the potential to effectively predict and monitor Quality‑of‑Life (QoL) indicators when designed with an appropriate methodology. This work aims to develop a comprehensive strategy for creating a specialized QoL‑monitoring application for head and neck cancer survivors.
First, the results of a literature review on mHealth services for cancer patients are presented. Then, the clinical study protocol is described in detail, within which patients are encouraged to:
Multicenter Randomized Trial for Assessing Quality of Life Using Non‑Invasive Intelligent Tools During Post‑Curative Follow‑up of Head and Neck Cancer Patients
- Head and Neck Medical Oncology Department, Fondazione Istituto di Ricovero e Cura a Carattere Scientifico Istituto Nazionale dei Tumori, Milan, Italy
- Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
- Oslo Center for Biostatistics and Epidemiology, University of Oslo, Oslo, Norway
- Oslo Center for Biostatistics and Epidemiology, Oslo University Hospital, Oslo, Norway
- Universidad Politécnica de Madrid-Life Supporting Technologies Research Group, ETSIT, Madrid, Spain
- Applied Research Division for Cognitive and Psychological Science, IEO, European Institute of Oncology IRCCS, Milan, Italy
- Department of Psychology, Educational Science and Human Movement (SPPEFF), University of Palermo, Palermo, Italy
- Information Technology Programme Management Office, DOTSOFT, Thessaloniki, Greece
- DeustoTech, Faculty of Engineering, Universidad de Deusto, Bilbao, Spain
- Institute of head and neck studies and Education, University of Birmingham, Birmingham, United Kingdom
- Division of Epidemiology and Health Care Research, JGU – Johannes Gutenberg University, Mainz, Germany
- Division of Oral and Maxillofacial Surgery – Bristol Dental Hospital, University of Bristol – Bristol Medical School, Bristol, United Kingdom
- Maxillo-Facial Surgery, Fondazione IRCCS Casa Sollievo della Sofferenza, San Giovanni Rotondo, Italy
- Maxillo-Facial Surgery, Interdisciplinary Department of Medicine, University of Bari “Aldo Moro”, Bari, Italy
- MultiMed Engineers srls, Parma, Italy
Head and neck cancer (HNC) survivors face significant physical, psychological, and socioeconomic burdens. Achieving cancer‑free survival while maintaining a high quality of life (QoL) is a central goal in the management of HNC patients, making lifelong follow‑up essential.
A particularly ambitious objective is to implement this follow‑up through advanced analysis of environmental, emotional, and behavioral data discreetly collected via mobile devices.
This clinical study aims to reduce the proportion of HNC survivors—who have completed treatment between 3 months and 10 years prior—experiencing clinically significant QoL deterioration during follow‑up, by leveraging non‑invasive tools such as patients’ mobile devices.
The Big Data for Quality of Life (BD4QoL) study is an international, multicenter, randomized (2:1), open‑label trial. Its primary endpoint is the evaluation of clinically significant QoL deterioration based on the EORTC QLQ‑C30 index over a 24‑month post‑treatment follow‑up period. The total study sample includes 420 patients.
Participants are randomly assigned either to the BD4QoL platform monitoring arm or to the standard clinical‑practice arm. The BD4QoL platform provides a comprehensive set of services for patient monitoring and empowerment through two core tools:
- a mobile application installed on participants’ smartphones, featuring a chatbot for e‑coaching,
- a Point‑of‑Care dashboard through which researchers manage patient data.
In both groups, participants complete QoL questionnaires at baseline and every six months, while also following the standard post‑treatment follow‑up schedule according to clinical practice.
Patients in the intervention arm (n = 280) gain access to the BD4QoL platform, whereas those in the control arm (n = 140) do not. Key inclusion criteria include completion of treatment for non‑metastatic HNC and use of an Android smartphone. Exclusion criteria include active treatment or the presence of synchronous cancers.
Digital Health for Supporting Head and Neck Cancer Survivors
- CONFERENCE: IEEE BHI-BSN 2021
IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI’21), jointly organised with the 17th IEEE-EMBS International Conference on Wearable and Implantable Body Sensor Networks (BSN’21)
This article presents the innovative capabilities that technological solutions can offer in supporting patients who have survived head and neck cancer.
The research was conducted within the framework of the BD4QoL project and focuses on leveraging digital‑health technologies to improve quality of life, monitoring, and overall support for patients during the post‑treatment period.
Smartphone‑Based Strategy for Monitoring Quality of Life in Head and Neck Cancer Survivors
- EAI PervasiveHealth 2022 – 16th EAI International Conference on Pervasive Computing Technologies for Healthcare
- apply self‑management practices through active symptom tracking,
- maintain healthy lifestyle habits,
- interact with embedded artificial intelligence that serves as an additional communication channel between patient and clinician,
- complete standardized QoL questionnaires via a web platform from home.
The challenges encountered, the selected non‑invasive data‑collection procedures, and the qualitative data derived from physical, social, and behavioral indicators together form a valuable set of guidelines and requirements for future research efforts focused on monitoring cancer survivors after treatment using portable IoT devices.
12–14 December 2022
Thessaloniki, Greece
Smartphone‑based IoT systems have the potential to effectively predict and monitor Quality‑of‑Life (QoL) indicators when designed with the appropriate methodology. This work aims to develop a comprehensive strategy for creating a specialized QoL‑monitoring application for head and neck cancer survivors.
First, the results of a literature review on mHealth services for cancer patients are presented. Then, the clinical study protocol is described in detail, within which patients are encouraged to:
Using Classification Techniques for Traffic Prediction in Smart Cities
- Conference paper – First Online: First Online: 29 May 2020 Part of the book series IFIP Advances in Information and Communication Technology (IFIPAICT, Volume 583)
Smart Cities are emerging as highly complex and interconnected ecosystems, offering intelligent services and innovative solutions to improve citizens’ daily lives. This work examines the use of classification techniques in Smart City projects, with a particular focus on traffic prediction.
Through a systematic literature review, the main thematic areas and methods used in smart‑city applications are analyzed, with special emphasis on data‑harvesting processes and data‑mining techniques.
The study investigates whether it is feasible to predict traffic load based on historical data and meteorological conditions. The results show that different predictive models can be developed using weather data, achieving varying levels of accuracy and effectiveness. These findings highlight the potential of applying AI and machine‑learning techniques to traffic‑management challenges in Smart City environments.
Smart Cities Data Classification for Predicting Electricity Consumption & Traffic
- Smart Cities are continuously evolving into highly complex and interconnected ecosystems, offering intelligent services and innovative solutions. These ecosystems generate and utilize large volumes of data from multiple sources, creating new challenges as well as opportunities for developing efficient smart‑city services and applications.
- the extent to which it is feasible to predict electricity consumption and traffic load based on historical data and meteorological conditions,
- which data characteristics are most suitable for prediction or for supporting decision‑making in matters of energy consumption.
The results show that in both cases, effective predictive models can be developed using weather data, highlighting the potential of applying AI and machine‑learning techniques in Smart City environments.
This work examines the connection between Data Mining techniques and Smart City projects through a systematic literature review that highlights the main thematic areas and methods applied. The research places particular emphasis on data‑harvesting processes and city‑data analysis using data‑mining techniques.
At the same time, two key research questions are explored:
Classification Methodology for Traffic Prediction: Highlighting the Impact of COVID‑19
- Journal: Integrated Computer-Aided Engineering, vol. 28, no. 4, pp. 417-435, 2021
This work presents an innovative classification methodology for day‑ahead traffic prediction. The research examines whether traffic conditions can be forecast based on meteorological factors, seasonality, time intervals, and mobility‑restriction measures related to the COVID‑19 pandemic.
Reliable predictive models are proposed that utilize smaller and more targeted datasets. Beyond feature‑selection processes, the methodology incorporates new attributes associated with COVID‑19 mobility restrictions, forming a new data model for traffic analysis.
The proposed approach investigates the optimal training subset for model development. The results showed that different models can be developed with varying levels of success, while the highest accuracy was achieved when all relevant features were combined and the proposed training subset was used.
The methodology demonstrated significant accuracy improvements compared to previously published approaches, highlighting the potential of applying AI and machine‑learning techniques to traffic prediction in Smart City environments.
Classification Methodology for Traffic Prediction: Highlighting the Impact of COVID‑19
- Issue title: Selection of papers from the 21st EANN (Engineering Applications of Neural Networks) and 16th AIAI (Artificial Intelligence Applications and Innovations) Joint International Conference
This work presents an innovative classification methodology for day‑ahead traffic prediction. The research examines whether traffic conditions can be forecast based on meteorological factors, seasonality, time intervals, and mobility‑restriction measures related to the COVID‑19 pandemic.
Reliable predictive models are proposed that utilize smaller and more targeted datasets. Beyond feature‑selection processes, the methodology incorporates new attributes associated with COVID‑19 mobility restrictions, forming a new data model for traffic analysis.
The proposed approach investigates the optimal training subset for model development. The results showed that different models can be developed with varying levels of success, while the highest accuracy was achieved when all relevant features were combined and the proposed training subset was used.
The methodology demonstrated significant accuracy improvements compared to previously published approaches, highlighting the potential of applying AI and machine‑learning techniques to traffic prediction in Smart City environments.
Game‑Based Smart System for Detecting Developmental Speech and Language Disorders in Early Childhood Communication: A Protocol Toward Digital Clinical Diagnostic Procedures
- Interactive Mobile Communication, Technologies and Learning Conference
- an interactive game‑based digital application for the child,
- an online data‑collection environment for parents and clinical specialists,
- a complete functional specification of the game‑based activities and the overall system architecture.
The proposed smart system has the potential to enhance digital‑healthcare processes for children’s communication skills, while contributing to positive social and economic impact in line with contemporary digital‑transformation trends in healthcare.
IMCL 2021: New Realities, Mobile Systems and Applications, pp. 559–568
Although international literature reports high prevalence rates of developmental speech and language disorders in children (3–17%), many cases remain undiagnosed, depriving children of timely and effective intervention.
The use of digital and mobile technologies in health and education creates new opportunities for monitoring, decision‑making, classification, and assessment processes. This study presents and documents a design and development protocol for a digital approach that supports screening and early‑detection procedures for developmental speech and language difficulties in early childhood communication.
The proposed solution leverages smart computing models, sensors, and early‑diagnosis indicators of linguistic and communication difficulties, and is developed in successive phases. The system design includes:
Co‑Designing User Experience in Location‑Based Games for a Museum Network: Involving Cultural‑Heritage Professionals and Local Communities
- Designing location‑based games (LBGs) for the cultural‑heritage sector requires the active participation of local communities and cultural‑heritage professionals to ensure authenticity, cultural relevance, and content validity.
- field visits,
- design workshops,
- field playtesting,
- field studies.
In parallel, the paper analyzes participatory‑design issues that emerged during the project, such as:
- the centrality and representativeness of participants,
- generating meaningful outcomes through collaborative sessions,
- co‑creating content through playtesting,
- the impact of the pandemic on participatory processes.
The work presents a comprehensive case of participatory and co‑design development of location‑based games for cultural heritage, emphasizing long‑term collaboration and continuous participant involvement in designing three different LBGs for a network of museums and cultural spaces.
This work presents a participatory and co‑design approach for developing LBGs that promote awareness and learning around intangible cultural heritage, specifically traditional crafts and artisanal techniques. The research was carried out within a long‑term project that covered the entire process—from initial awareness‑raising to final implementation.
Following a design‑thinking methodology, the study outlines the participatory methods used to engage cultural‑heritage professionals, local communities, and visitors of museums and cultural sites. These methods include:
Personalization Framework for Cultural Routes: Case Study of the TRACCE Project
- Recent developments in digital‑storytelling technologies within cultural tourism have strengthened new trends toward more comfortable, interactive, and personalized travel experiences. The emergence of recommendation systems and personalization techniques creates new opportunities for enhancing the experience of cultural‑heritage visitors.
This work presents an innovative narrative system for cultural routes, which combines users’ multimedia‑content preferences with different typologies of cultural tourists.
As a research foundation, the McKercher typology is employed, categorizing cultural‑heritage visitors into five main groups. Using a specially designed questionnaire, an enriched typology of cultural tourists is developed, adding three additional categories based on users’ preferences regarding multimedia content.
The proposed approach contributes to the creation of more personalized and interactive cultural experiences, enhancing visitor engagement and interaction within cultural routes and heritage environments.
Redefining Smart Tourism through Smart Gamified and Gaming Applications in Greece
- World Academy of Science, Engineering and Technology – International Journal of Social and Business Sciences
- “Hologrammatic Corfu” for the old town of Corfu,
- “BRENDA” for the Kilkis region.
Although the two systems were designed independently and respond to different needs, they are based on a shared gaming and gamification methodology.
Hologrammatic Corfu was designed to support comprehensive exploration of the city before, during, and after the trip, using transmedia content such as:
- photographs,
- 360° videos,
- augmented reality,
- hologrammatic videos.
In parallel, statistical analysis of point‑of‑interest visitation is used to dynamically redirect visitors, enhancing sustainability, accessibility, and balanced management of tourist flows.
BRENDA was designed to promote gastronomic and historical tourism, using serious‑gaming and gamification mechanisms that connect local businesses with cultural points of interest. This approach encourages active participation from both local communities and visitors, fostering sustainable and interactive tourism experiences.
Finally, the work highlights the potential for reusing gaming mechanisms and transmedia elements in new areas of interest, contributing to the transformation of destinations into modern smart destinations.
Conference Title: ICSTDM 2021 – International Conference on Smart Tourism and Destination Management
Conference Location: Istanbul, Turkey Conference Dates: 28–29 January 2021
Vol. 15, No. 01, 2021
Smart technologies are increasingly used to enhance the travel experience and shape a destination’s image through digital applications and social‑media platforms. This work presents the design and implementation of smart‑management applications that promote culture, sustainability, and accessibility in two different tourism destinations in Greece.
The first case concerns Corfu, a highly popular destination and UNESCO World Heritage Site, where increased visitor numbers create challenges in experience organization and infrastructure management. The second case focuses on Kilkis, a destination with low tourist traffic and high seasonality, where traditional approaches to promoting culture and local experiences have not produced the desired results.
To address these challenges, two different systems were developed:
Participatory Urban Planning through an Online WebGIS Platform: Functions and Tools
- This work presents an online WebGIS platform for participatory urban planning, focusing on its objectives, functionalities, and tools within the context of the ppCITY research project.
- the creation of projects and sub‑projects aligned with the stages of participatory urban planning,
- the use of libraries of proposed solutions,
- the combination of user profiling and opinion weighting,
- the management and analysis of geospatial data through WebGIS tools.
The primary goal of the platform is to highlight the essential characteristics of public space and to organize processes of meaningful and equitable citizen participation in urban planning.
This research approach also aims to create an open community of users, experts, planners, and developers around an open‑source ppWebGIS software environment, fostering collaboration and participatory innovation in the fields of Smart Cities and urban planning.
The development of Information and Communication Technologies (ICT), and particularly WebGIS tools, has significantly strengthened public‑participation processes, expanding the ability of diverse stakeholder groups to engage in urban planning and decision‑making.
Technically, the platform is based on open‑source ppGIS solutions and integrates innovative online tools that support the core methodological stages of participatory planning, leveraging real spatial and urban challenges.
The system enables:
