Sunday, June 16, 2024

AI-based platform offers personalised support to people with obesity

Must read

Using cutting-edge AI techniques, the Stop Obesity Platform features a chatbot that tailors recommendations on nutrition and exercise according to people’s health data and emotions.


Overweight and obesity affect 59 % of adults and almost one in three children in Europe, according to the WHO European Regional Obesity Report 2022. Obesity has become a societal challenge, but the STOP project, undertaken with the support of the Marie Skłodowska-Curie Actions programme, wants to help change that through AI.

Coordinated by FTK, Research Institute for Telecommunication and Cooperation in Germany, the project developed an innovative platform that gathers and analyses health data from people with obesity to provide them with information and guidance related to nutrition and exercises. It also helps mitigate the growing costs of obesity and related health issues, such as heart diseases, diabetes, liver disease, gallstones, cancer and dementia.

“One of the key features of the Stop Obesity Platform is its ability to import users’ fitness data from wearable devices,” says Binh Vu, project coordinator. “To harness the power of this rich data set, machine learning algorithms are employed to analyse and derive valuable insights from the collected information.”

The data collected includes people’s activities and exercises, along with nutritional, biomedical and physiological information, used to create a comprehensive health profile for each user. Then, the algorithms analyse patterns, correlations and trends within the data to provide a deeper understanding of users’ health behaviours and needs. “This analysis serves as the foundation for the platform’s chatbot functionality, which offers a unique way to receive personalised health information and guidance. Together with healthcare professionals, the chatbot can suggest better nutrition and exercises to patients,” describes Vu.

AI tools create an empathetic chatbot

The integrated chatbot made data collection easier and more frequent in the STOP platform because users could engage with it at their convenience. Additionally, it was developed to observe and analyse the conversations, adapting them to the patient’s mood. AI techniques played a pivotal role in this, showing its significant value in virtual psychological counselling.

Natural language processing, a subfield of AI that helps computers understand, interpret and manipulate human language by transforming information into content, was used to enable the chatbot to effectively understand user input and respond to it accordingly. The research team worked with deep learning models, like recurring neural networks and transformers, to handle the complexity of natural language understanding and generation tasks.

Sentiment analysis was fundamental to allow the chatbot to assess the users’ emotions during conversations and provide more empathetic and personalised support. “By utilising advanced deep learning models like convolutional neural networks and long short-term memory networks, we were able to accurately analyse the sentiment expressed in user input,” explains Haithem Afli, lead researcher of the AI chatbot.

The project team also implemented personality models to the chatbot’s conversational style to make interactions more engaging and user centric. “The advancements in the used AI techniques made a significant difference in providing effective support and guidance to individuals with obesity,” Afli points out.

Gamification for continuous results

The Stop Obesity Platform also takes advantage of gamification to encourage people with obesity to continue practicing healthier habits. For example, users can earn virtual points when they reach their fitness goals, learn about nutrition through mini games and participate in challenges to improve their activities’ progress. The idea is to provide gratification and tangible measures of success to make the behaviour change more enjoyable.


STOP, obesity, chatbot, AI, Stop Obesity Platform, health data, deep learning, gamification, machine learning, healthcare, GPT, NLP, natural language processing, neural networks, convolutional neural networks

Project Information


Grant agreement ID:

Open in new window

Project closed

EC signature date
24 September 2018

Start date
1 March 2019

End date
28 February 2023

Funded under

  • EXCELLENT SCIENCE – Marie Skłodowska-Curie Actions

Total cost

929 200,00

EU contribution
€ 929 200,00

Coordinated by


This project is featured in…

Latest article