Stress Analytics (SEAL) Project

Stress is a natural state of physical and mental tension induced by common external and environmental factors, such as challenging personal and work situations that each of us have to face every day. Driving our cars in traffic, dealing with personal issues, job-related issues (workers), school tests (students), taking critical decisions (doctors, financial professionals); all of these are situations require acute psycho-physical changes. In turn, that process may induce anxiety and/or other physical symptoms (such as hypertension, headache, gastrointestinal diseases, etc…) in some people. Present scientific understanding of stress-related health problems supports the view that, for improving their stress resilience, people should be able, first to identify potentially stressful situations in their lives, and to adopt appropriate stress coping strategies and life style. However, to achieve such goals this field of research still needs to collect, interpret and react to ecological behavioral data and metadata from our personal and social life.



The main goal of the SEAL Project is the collection and analysis of dyadic Spoken Conversations’ overt and covert signals, namely speech and physiological signals, in a naturalistic work environment (e.g. call center). The following signals and metadata will be recorded and collected as part of the project:

a) Spoken Conversations (operator and customer side)

b) Physiological (Operator Side) Stream

c) Evaluation of employees’ work-related stress.

The ultimate goal of the SEAL project is to investigate the stress state and its observables in naturalistic settings. We plan to correlate it with diverse overt/covert signals such as GSR, heart rate, spoken language, behavioral markers and understand their interplay as well.  Modeling such signals will account for human-expert annotations (both self-assessment and expert evaluation). The SEAL project will bring groundbreaking signal, corpora and models to advance technology that can impact positively people’s life (workers’ and businesses’).

Research Team

Giuseppe Riccardi, Arindam Ghosh, Morena Danieli