IST_652_Final_Project_Proposal.docx
pdf
keyboard_arrow_up
School
Syracuse University *
*We aren’t endorsed by this school
Course
652
Subject
Industrial Engineering
Date
Dec 6, 2023
Type
Pages
3
Uploaded by DoctorMonkey21193
Final Project Proposal
IST652 - Scripting for Data Analysis
Team Members: Akshitha Reddy, Ramesh Lakshman, Reshma Kalla
1.
Choose whether to work individually or to work in a team of 2-3 people. If you wish to work in
a team, specify the people that you have talked with to form a team.
We will be working as a team of 3 on the final project for IST 652 - Scripting for Data Analysis
Team members: Akshitha Reddy, Ramesh Lakshman, Reshma Kalla
2.
Pick a topic of investigation. [1 point]
Topic
Spotify viewership based on Factors of Song and User Playback (e.g. 1920- 2020).
Links to datasets
https://www.kaggle.com/code/vatsalmavani/music-recommendation-system-using-spotify-dataset/input
Brief Explanation
First and foremost, we want to learn more about how consumers listen to music on Spotify. This
data can be used to enhance users' Spotify experiences, such as recommending new music that
they are likely to appreciate. For example, if a user frequently listens to songs by a certain artist,
Spotify may suggest other songs by that artist or comparable artists.
Second, we want to assist artists and record labels in better understanding their target
audiences. This data can be utilized to develop more focused marketing efforts and to generate
music that is more likely to succeed. For example, if a musician understands that their fans are
more likely to listen to music on Spotify on weekends, they could plan their new music releases
accordingly.
Here is a set of initial questions once the data has been collected :
●
Are there any trends in viewership by genre, language, or other factors?
●
Is there a correlation between viewership and other factors such as social media
following, critical acclaim, or commercial success?
Following the initial data analysis, we may have additional inquiries and find additional
prospective development activities:
●
Develop data visualization to help users explore and understand Spotify viewership data.
This tool could allow users to filter and sort data by different criteria, to create custom
visualizations, and to share their findings with others.
●
Develop a machine learning model to predict future viewership for songs and
artists.(Future Possibility)
3.
Pick the data sets you plan on using. [1 point]
Currently, we are planning on using the Spotify dataset available on Kaggle. The dataset includes
information related to music recommendation systems using Spotify, which aligns with our goal
of understanding Spotify viewership based on factors of songs and user playback. The dataset is
categorized into different datasets with respect to genre, artists, and year, this allows us to
explore trends in viewership based on these factors. We are also in a hunt for an exclusive
dataset that provides us with the viewership of Spotify so that we can connect our current
dataset to the viewership dataset and figure out the trends in which consumer’s choices are
changing. Based on these results we will further try to bring in predictive analysis on the same.
These predictive analysis will be further used for strategic marketing by the artists and music
labels.
4.
Pick several (2-3) possible methods of data acquisition and analysis.
The spotify data we have is categorized into different datasets with respect to genre, artists and
year. We might merge these datasets using year as a common parameter to increase the
efficiency of the model.
Since there are several factors that affect user experience like danceability, acousticness,
duration etc., we should come up with a standard unit of analysis for accurate results.
We may use several visualizations like histograms, scatterplot matrices to analyze and interpret
the relationship between dependent and independent variables i.e., user experience vs.
Acousticness, danceability etc.,
It is likely that we use linear and logistic regression models to deduce which variable has the
highest impact on the user experience and how it is changing over the years.
We might have to normalize certain variables because it is difficult to interpret logit for large
numbers.
5.
Identify potential development tasks (2-3) and whether you believe you’ll need additional
guidance in achieving results. [1.5 points]:
1)
Could use some insight/help on this but would be curious if there would be a way to write a
script that would help us clean datasets as we are working with them in MBs and manually
wouldn’t be effective.
2)
We will have to develop a script that will group Songs into the same group based on categories
and song factors such as synergy, loudness etc. This will require segregation methodologies that
would help us visualize and better understand data at hand.
3)
Once the songs are categories based on different factors, comparisons can be done using
population dataset.
If the foregoing isn't enough, it could be interesting to broaden this project to include one of the
following.
It could be interesting to develop a predictive model based on Spotify song viewing, song quality
factors, drop days, and so on. However, most of this will be determined by our preliminary results.
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
- Access to all documents
- Unlimited textbook solutions
- 24/7 expert homework help