Assessment of Lifestyle and Mental Health: Case Study of the FST Beni Mellal

摘要

Lifestyle habits are defined as behaviors of a sustainable nature which are based on a set of elements incorporating cultural heritage, social relations, geographic and socio-economic circumstances as well as personality. Mental health encompasses the promotion of well-being, the prevention of mental disorders, and the treatment and rehabilitation of people with these disorders. In order to address this issue, we propose a solution which consists of the development of an extended autonomous computer model for large textual data. This model will make it possible to give a psychological, emotional or even a lifestyle character from tweets or a web forum. So we turned to the notions of sentiment analysis and Text Mining using Deep Learning. This work (which will be limited to a Moroccan context) concerns the development of a computer model that allows to determine the habits of life and the Health of the students of the Faculty of Sciences and Technologies at the Sultan Moulay Slimane university in Beni Mellal. We started by developing a script to retrieve posts made by students from a Facebook group. The choice of Facebook and not Twitter is due to the fact that the twitter community among the students is relatively small. Afterwards, we built our deep learning model and we tested it with data from twitter comprising of thirteen (13) classes (anger, joy, sadness, disgust etc.). We also submitted these textual data to automatic learning algorithms (naive Bayesian, K nearest neighbors).