UT Arlington Libraries City of Searches

The City of Searches visualization project was inspired by a NISO webinar in September 2018 called Using Analytics to Extract Value from the Library’s Data.  Steven Braun from Northeastern University Libraries shared a visualization that he had created to highlight the Libraries' catalog searches for the previous year.   Pushyahass Reddy Obulapu, a CSE Masters degree student working as a GRA in Libraries Assessment took on the assignment of developing a visualization for the Summon searches for UTA Libraries for the 2017-18 academic year.  

His notes and comments about the project are included below, as well as the resulting visualization.  

Data Visualization of Search Terms at UT Arlington Libraries

• Students at The University of Texas Arlington Libraries have searched for a wide range of topics at the Libraries website. All the search terms are clustered by category and are visualized as a city, with each category represented as a building block.
• Each individual building block is scaled dynamically depending upon the frequency of the search category. (Large "buildings"represent more frequently searched topics, small  ones represent less frequently searched topics.)
• Each sub block represents the most commonly search term in a particular category.
• This data visualization tells fascinating stories about what students are interested in or searching for at UTA Libraries.
• The raw search terms are preprocessed and clustered to different categories using pandas and numpy Python Libraries.
• The clustered data is dynamically visualized using D3.js, a JavaScript Library.
• Before the start of this project, being a Computer Science graduate student, I had a basic understanding of the D3.js Library and had good hands on experience with python. This project gave me a challenging environment and helped me to apply my programming skills in a broader perspective and in learning D3.js to perform interactive data visualizations more efficiently.