5 Quirky Data Science Projects to Impress

Develop unique yet standing-out data science projects to improve your data portfolio.



5 Quirky Data Science Projects to Impress
Image by Author | Ideogram

 

Data science is a field many try to enter, but fail attempting to do so. It’s not because they are incapable; it’s that they are not standing out from the crowd. There are so many applicants for small pools of job openings that competition becomes fierce. That’s why data science applicants must be creative with their projects.

You might not necessarily find a quirky project in the business in general, but it could stand out and showcase your capability that can’t be shown in basic projects. To help, this article will explore 5 ideas for quirky data science projects you can develop to impress everyone.

What are these projects? Let’s get into it.

 

1. AI-Powered Dungeons & Dragons Dungeon Master

 
I am an avid gamer and have always loved great gaming-related projects. As such, the first project idea is developing a dungeon master for Dungeons & Dragons (DnD) with AI. DnD is a complex game with many mechanisms, and developing this project can be a fun and informative data science project.

The project concept could be an LLM-powered dungeon master, such as using a GPT-based model for text generation. The dungeon master controls the game, so the tasks should be generating a story, managing non-player characters, interacting with players and reacting to their decisions, providing items and enemies, and other related tasks. The main logic can be built using Python, and a simple interface created with a framework such as Streamlit can be used to enhance the project.

Try to structure what the dungeon master can do and provide each of these features in your project.

 

2. Art Style Transfer Model

 
There are many artistic styles out there, and it would be an interesting project to develop a model that can transfer one art style to another. For example, consider having a simple drawing that you want to turn it into Picasso style, or you have an anime style picture that you would like to transfer into a Van Gogh painting style.

This project would involve neural network model implementations for neural style transfer using a framework such as PyTorch or TensorFlow. You could use pre-trained models such as VGG19 and a library such as OpenCV for image processing. Matplotlib and Streamlit could be used as the front end for visualizations.

The project may seem simple, but it can showcase your ability to work with popular frameworks and libraries, and to understand neural network models.

 

3. Historical Data Visualization

 
Many people, including myself, love great historical stories. Taking this into account, what better project than one using historical data data to tell a story visually? This visual aspect would make the project even more interesting than telling the same story with text.

You could collect data from open source datasets like those available from Kaggle or numerous governmental archives, to name but a few sources. For the visualization technique, we could try to create a historical event timeline progression, interactive geographical map over time, or simple heatmap to show event intensity of a phenomena through time. Python library tools such as Seaborn or Plotly could be used to accomplish these goals. A dashabord could also be implemented using tools such as Power BI or Tableau.

This kind of project is interesting on its face, but also has depth and complexity that would impress the others.

 

4. Weather-Based Outfit Recommender

 
This project might seem simple, recommending outfits based on the weather. How hard could it be, right? However, the project can become much more complex once we consider the data we wold actually need and the model development process.

The recommendation must be based on weather forecast data, the user's outfit preferences, and seasonal fashion trends. Using a weather service API, a clothing dataset's metadata, and collected user style preferences, we could build a suitable classification model with a decision tree for each potential weather outcome in order to get the recommendations. Try to have a simple front-end using Streamlit to visualize the result.

Are you worried that it would not be impressive enough? With a little creativity, a recommender could easily be extended to other topics in other domains.

 

5. Time Capsule Sentiment Analyzer

 
Sentiment analysis is a common project often used to understand the nuance of some particular aspect of business. But how can we extend the project to analyze the progression of sentiment over a period of time? In this proposed project, we will use a sentiment analyzer to see how our personal writings or social media posts evolve over time.

Using the text data collected between certain periods, we can track how the text's mood, sentiment, writing style, subjectivity, as well as other characteristics. The progress can give you insight into personal growth and the changes in style over time. Alternatively, you could use a business' social media data to see the progression of the business sentiment over time.

This project could also involve multiple modelling approaches. Not only could sentiment analysis be employed, but time series analysis could also be used to gain trend insight, or topic modelling to see the prevelant topics in a certain time period. You could then try to visualize these findings using simple dashboards like Power BI or Tableau to make them more interesting.

 

Conclusion

 
In this article, we have explored 5 quirky data science projects meant to impress. By implementing any of these, or using them as inspiration for other projects, you can go a long way toward standing out in the sea of data science applicants and showcase your skills in different ways.
 
 

Cornellius Yudha Wijaya is a data science assistant manager and data writer. While working full-time at Allianz Indonesia, he loves to share Python and data tips via social media and writing media. Cornellius writes on a variety of AI and machine learning topics.





Our Top 3 Partner Recommendations



1. Best VPN for Engineers - 3 Months Free - Stay secure online with a free trial

2. Best Project Management Tool for Tech Teams - Boost team efficiency today

4. Best Password Management Tool for Tech Teams - zero-trust and zero-knowledge security