If a picture is worth a thousand words, then the information-rich images of data visualizations convey thousands of points of information.
That idea was aptly illustrated by members of undergraduate and graduate classes who took part in the School of Information Studies (iSchool) Fall Poster Day Session. Several dozen projects illustrating research questions—and answers that the data assessments revealed—were on display for iSchool faculty, staff and fellow students. Representing dozens of final class projects, the posters showed data findings formatted as bar charts, line graphs, scatter plots, heat maps and the many other ways that students have learned to present information visually this semester.
For their assignments, students were asked to research an issue or problem then show the answers or solutions they discovered trough programming languages and tech tools in visual ways. Poster topics ranged from an analysis of gun violence incidents in America, to attrition rates of employees, to differences between male and female and summer and winter Olympic athlete participation, to how to gauge a satisfactory chocolate bar.
One student team used LangDetect python library and VADER, a lexicon and rule-based sentiment tool, to analyze four months of tweets surrounding the #MeToo sexual assault reporting movement. Through data extraction, reprocessing and sentiment analysis processes, they studied the polarity of statements on Twitter and trends of social media engagement. Binyue Wang, Deqian Kong, Junqing Xa and Linwanting Wang, students in IST 664, a natural language processing course, examined several aspects and trends of the issue. One of the lessons they learned, they said, was that a deeper dive into the data was necessary to more thoroughly flesh out the issue. According to Kong, they were able to tally Twitter remarks into negative or positive comments, but those sentiments alone didn’t yield in-depth information about the perspectives and attitudes of those making the statements.
Senior James Lu mapped statistics regarding the Olympic games over a span of 20 years. He compared rates of male vs. female athlete and summer vs. winter-games participation, as well as noting finer details such as the countries of athletes and their body mass indexes and competition ages. He worked with an initial dataset containing 270,000 rows and 15 columns, and his poster used a final dataset of 87,444 rows and 10 columns of information.
Though Lu is an Olympics enthusiast who envisions a future in data analytics work and perhaps doctoral studies in that area, he was still surprised by some of the findings. “It was shocking to see a decrease in events for men in summer and a significant spike for both men and women in the winter games, especially since 2010 and beyond,” he added. His experience convinced him how helpful and applicable data visualizations can be in the business world, he said “particularly when you have so much data but you want to present it concisely to your boss.”
More than interesting pictures results from shaping data into images, said student Rucha Kadam, who mapped statistics related to gun-violence incidents in the U.S. for IST 719, a data visualization class. Using data obtained from Kaggle, an open repository of data sets, and four years of incident reporting, Kadam said certain trends, as well as specific questions, stood out. Those insights make it feasible to consider how changes in policies and regulations might reduce levels of gun violence, she noted, such as “looking at suicide and depression rates along with peoples’ possession of guns,” and “looking into age limits for the issuance of gun licenses.”
For their IST 654 Systems and Analysis Design project, Paul Rayi and Nidhi Shivak conveyed a multi-faceted technology concept they developed to meet the need to match employer jobs with employee skills. Called “Bridging the Gap between Skills and Jobs,” their system connects employees and employers, suggests specific skills potential employees need to land certain jobs and recommends sources where they can learn those missing abilities. The system also accounts for changing industry needs and working trends and offers career counseling assistance. Rayi, who plans to undertake a Ph.D. in information science, said he was pleased at the reception their system design was receiving from the Poster Day audience: “I like the reviews so far.”
The range of posters fall students presented showed the diversity of research questions and the many different ways discovered data can be displayed. Among projects topics were: “Dining Health in NYC,” Suicide Statistics and Analysis,” “IT Employee Mental Health Survey Analysis,” “World Happiness Index,” “Los Angeles Metro Bike Share—Who’s Using it and Why,” “Employee Attrition Analytics,” “Terrorist Attacks from 1970 to 2017,” “Volcanic Eruptions,” “Chocolate Bar Ratings,” and many others.