// new media class

The project was inspired by the emergence of fake artists on Spotify. There are real people behind fake artists, but they do not exist outside of the streaming platforms. They have neither a social media profile nor a fanbase. This is because they mainly produce music for so-called mood playlists – and generate millions of streams with it. Oftentimes, many of the fake artists have the same producers and companies hiding behind them.

Such fake artists will emerge in large numbers, the root of the problem is the increasingly passive nature of music consumption. People will often ask Alexa, or some other digital assistant, to find background music for a specific task—studying, workout, housework, relaxing, etc. Or they will rely on a pre-curated playlist for that purpose. They don’t pay close attention to the artists or song titles, and this is what creates an opportunity for abuse.

My initial expectation is to strengthen the connections between music and what we're doing right now. In the meantime, I want to reduce the amount of work users have to do to search and select soundtracks in Spotify and then present the corresponding image on the screen.

This project is divided into 2 parts.

Part 1:The first step is for the user to enter the link of the website they are viewing into the site (ideally this is the only thing they need to do). The second part is divided into two parts. The first part is to capture and extract keywords from the textual content of the target website by using beautiful-soup, and then count the number of times these words appear. If words in the same category occur more than five times in total, output the parent words according to the pre-set categories. Then use selenium to automatically jump to enter the keywords into the Spotify search box and play the first playlist.

Part 2:Analysis of sentiment in the text can help determine the options and affective intent of writers, as well as their attitudes, evaluations, and inclinations concerning various topics. Work on sentiment analysis has typically focused on recognizing valance - positive or negative orientation. In this part of my work, i want to explore the task of identifying expressions of emotion in text. And present the dynamic image on the background of the website and fade out.