Using Python-based face recognition technology, this work aims to forecast characteristics including age, gender, and moods of individuals shown in video uploaded on the YouTube platform from 2017 to 2021. This study particularly makes use of this method on individuals engaged in YouTube street interviews, focusing on issues related to the unemployment and economy.

The concordance and distribution of emotional variables alongside demographic variables such as age and gender were investigated using the "correspondence" analysis method, a multivariate technique that visualizes relationships among variables and categorical cross-relationships through graphical mapping.

Data concerning the emotional variance from every year between 2017 and 2021 were gathered. Data were especially separated, and unhappiness rates were computed as percentages from the gathered data. The study looked at the relationship between the unemployment rate for the particular years and the computed degree of discontent. Non-parametric statistical methods were seen more suitable in this phase of the research as the low data volume made normalcy testing impossible.

The value of facial recognition systems in demographic prediction and emotional evaluation is underlined by this work. Furthermore, it emphasizes the possible value of matching emotional forecasts with unemployment data in order to assess how economic changes affect personal emotional responses.

This study emphasizes how improvements in artificial intelligence and data analysis could improve the understanding and answers for society issues. Therefore, it could motivate similar future study with great benefits in marketing, economics, and social sciences as well as highlight developments in data analysis and artificial intelligence.