ABSTRACT
In the era of Covid 19 pandemic, governments have imposed nationwide lockdowns which make a huge change to people daily routines. This last affect indirectly on the well-being of people’s mental health, especially the vulnerable population. And due to social media, many conversations about these phenomena occur online, especially those related to people’s emotions. Then the field of sentiment analysis is requested. In this paper, we aimed to extract correlations within this epidemic and its psychologic effects. In fact, our goal is to extract features that may improve sentiment analysis accuracy which is a crucial step to fulfill the main objective of our research: developing an intelligent recommendation system that will benefit persons, through a positive accompaniment and the early alert, in case of complex situations as mental illness. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.