From the Oven to the NYC Subway
By Saummya and Abhiram Reddy Kadimetla
I have always been a big fan of New York City, with one of my biggest goals in life living in midtown Manhattan. The one thing that pulls me towards the Big Apple is the scores of pizza restaurants, each one unique and serving the coveted one slice of greasy goodness. While New Yorkers cannot take credit for inventing pizza, they are the true connoisseurs of the one cheesy slice of pizza with a generous amount of marinara sauce. No other city in the world has come closer to replicating the slice, some even crediting the differentiating factor of the Big Apple pizza to the minerals in the water of the city.
The stats on the Big Apple
There are approximately 9,000 pizza places in the city. Most New Yorkers have their one hole in the wall place that they like to frequent. When I first visited New York, I heard and read contrasting reviews from true blood New Yorkers, which prompted me to go on a quest to find which ones are the best.
In 2022, a whopping 56 million number of tourists are projected to visit New York City. Their number one preferred mode of transportation? The subway! The 24 hour NYC subway station network consists of 36 lines with 472 stations. You can imagine how difficult it must be for a rookie or a tourist to navigate the subway. So how would a tourist find the perfect pizza slice near them while navigating the subway, without having to comb through Yelp? This is the question we set out to answer through our visualization project.
Finding Inspiration
We combed through the internet to find similar works. We ended up discovering two such works which heavily influenced our project. The data visualization technique takes help from two major works on this topic, “Exploration of Pizza restaurants in New York City” and “New York, Sliced”. Both of these works contain four datasets in total which have data from reviews by critics and reviews on websites like Yelp and Barstool. These datasets help visualize data in various ways ranging from location specific, Likert scales and ranking top 20 spots in the city. However, the visualizations in the first work are visually unappealing and do not tell a story. Similarly, the visualizations in the second work tell a story but contain a lot of heat maps which are informative but not easily interpreted. It focuses too much on location. This might throw off people who are not familiar with the areas of Manhattan.
We also looked through some research papers and other literature related to our project. The researchers in the paper “A survey on information visualization: recent advances and challenges” discuss unusual or uncommon data visualizations and their pros and cons. We took inspiration from this paper in terms of the unusual cartographic visualizations that they showcased. The one visualization that particularly interested us was a typographical map of the streets of Chicago. The type of visualization that it falls under is a typographical map that merges text and data. This was one of our goals for our map based visualization.
Devising Our Approach
Our approach to this project was to use these existing datasets and create visualizations which could be easily interpreted by tourists and first timers. It would focus more on storytelling and creating visually appealing and informative data visualizations. A further idea of ours was also to visualize this data in AR which could recognize the location and provide the review of the place. We also wanted to leverage the existing heatmaps and create a better navigation system, integrated into the visualizations.
Giving Life to the Slice
We knew that we wanted to somehow combine subway navigation with the pizza places. Initially we thought of a dashboard design to provide a map and navigation based visualization. However, this was not innovative enough and already existed on popular review websites. We went back to our literature review and found the one key thing that we were missing from our datasets: The lines on the NYC subway. The subway network visualization is one of the oldest visualizations in the world. We decided to make this visualization a part of our final project and sketched a low fidelity version of the idea
Fighting for the Final Slice
We began by putting the datasets on Tableau and joining them to create simple bar chart visualizations. The first visualization consists of a bar chart with the pizza restaurants and at what subway stations they are located at. The Y axis is the number of stars that the restaurant has from 0 to 5 with a step of 0.5. Note that all subway stations in the city do not have pizza restaurants near them.
This visualization has a MAX function applied to it, where only the restaurants with maximum stars at each subway station are displayed. The restaurants are sorted according to each subway station.
Of course, we did not stop here. As designers we wanted to make our visualization user friendly and visually appealing, so we decided to make a subway map with different color coding for the lines. Then we put a pizza icon for each subway station that had restaurants with more than 3.5 stars near it.
We also wanted to make it navigable. We implemented an interaction where if a user clicked on a line, it would get highlighted and the restaurants on that line would pop up. The user can click on any restaurant and see how to reach it. A demo of this interaction is below:
We chose to use Adobe Dimension and Adobe Aero for the Augmented Reality visualization. We designed the 3D assets on Dimension which consisted of a block containing the restaurant’s name, number of stars and a Google Map. Then we used an image target on Aero to spawn the blocks over the image. The image was a representation of the subway signs that are commonly seen all over NYC outside the subway stations. This is a demo of the AR visualization:
Future Scope
The AR visualization was a bit difficult to understand for our users. We plan to include an onboarding and a help feature in the app to make it easier to use. Furthermore, the lines on the map visualization are also intertwined and difficult to navigate on handheld devices. We want to optimize this visualization for mobile and other such devices.
Acknowledgments
We wanted to thank Professor Niklas Elmqvist, our TA Mofe Barrow and our peers in INST760 Data Visualization class.
