Electric Vehicle Infrastructure: Week 3

Hello week three! Last week, team EV kicked off the project by doing an exploratory data analysis to become more familiar with the dataset that we are working with. We found several interesting results and have posted them here so readers can take a look!

This week, we started to dive into our first two subprojects: sentiment analysis of reviews and categorization of reviews based on topic. For sentiment analysis, we’ve successfully built a Support Vector Machine model that is currently able to predict whether a review has positive or negative sentiment at around 84% accuracy when compared to hand labeled reviews. This model is currently performing better than the previously used model, which was created using Microsoft Azure Machine Learning Studio, but we are still looking for ways to improve it further.

For our subproject on review categorization, there is significantly more prep work that must be done before analysis. Before we do data analysis, we must categorize over 140,000 reviews! Our plan is to first build a training set on 14,000 reviews, and then utilize machine learning techniques to classify the majority of the reviews. Even so, 14,000 is a lot of reviews, so the EV team is going to crowdsource categorizing a portion of them to Amazon’s Mechanical Turk. While this will ultimately speed up our process, we have to write an IRB protocol and set up a proper environment to collect valid training data from MTurk. These two steps, writing the IRB and setting up MTurk, are where we have focused our work on subproject two this week. At the end of next week, we’re hoping to have the all of the materials set up to start requesting Turkers to complete our tasks.

Electric Vehicle Infrastructure

This summer, our team is assessing the existing electric vehicle (EV) infrastructure in the United States to better understand if it is serving the public properly. To do this, we’re analyzing charging station review data obtained from a popular mobile app that works like a Yelp-like review system for EV charging stations. We are particularly interested in exploring the differences in quality and functioning between public and private charging stations as well as analyzing the different topics of discussion within the reviews and the trends within them. Once we understand how everyday users are interacting with EV charging stations, policy recommendations on infrastructure can be made to effectively support the growing number of electric vehicle owners.

Our project is split into three possible subprojects:

Project 1: Sentiment analysis of reviews.

Here, we are going to use an existing training data set to classify all of the popular mobile app’s EV station reviews by sentiment to discover general sentiment regarding EV charging infrastructure.

Project 2: Categorization of reviews

Next, we are going to categorize all the reviews according to the subjects of the reviews. This will be much more labor intensive because we will need to create a training data set before we can analyze the major concerns of EV users.

Project 3: Predicting station failure

In our final phase, we will analyze real-time data from EV charging stations to create a model to predict when an EV station will break down.

By the end of the summer, we hope to complete all three projects, but we know that we are operating under a strict time constraint, so accomplishing all might not be possible. Despite this, we are excited to tackle all of this this summer!

Team EV with our mentor, Professor Omar Asensio and graduate student Catharina Hollauer exploring Atlanta and getting to know one another at a Braves baseball game.