By Jayasree Jaganatha, Renaissance Computing Institute (RENCI) –
Notes, takeaways, and resources from our fifth event (featuring Snigdha Chaturvedi, David Gotz, and Danielle Szafir) –
On June 23, 2022, Carolina Data Science Now hosted the fifth iteration of its seminar series, which will strive to link the people, research, and resources by highlighting the many faces of data science across our campus, and demonstrating the integral role data science can play in all disciplines.
This Month’s Speakers
The theme of the event was “Data Science: More Than Numbers” The event was moderated by Mariana Olvera-Cravioto, and the speakers were selected to provide insight on how data science is used as a foundational anchor in computer science and statistics projects.
David Gotz | School of Information and Library Science
David Gotz spoke on various efforts from the “The Visual Analytics and Communication Lab,” which conducts research at the intersection of interactive visualization, human-computer interaction, and machine learning. Read more about the lightning talk here.
Danielle Szafir | Computer Science
Danielle Szafir presented “Making Data Visible: Cognitively-Driven Data Visualization @ The VisuaLab,” which provided an overview of projects at The VisuaLab, comprised of researchers working to make data analysis more efficient by studying, understanding, and predicting how people will work with and interpret data.
Read more about the lightning talk here.
Snigdha Chaturvedi | Computer Science
Snigdha Chaturvedi discussed “Data Storytelling through Natural Language Generation,” and presented projects that implement natural language processing (NLP) to convert computer data to human text. Read more about the lightning talk here.
Questions for the speakers were collected through an EasyRetro board. Topics during this section included:
- Exploratory avenues for getting involved and learning more about data science, including choosing tools depending on the expectations and quality of your data and utilizing tools that have been created for similar purposes.
- The variety of data science tools and services used across campus, including Tableau, R Studio, Jupyter Notebooks, and Excel.
- Whether monitoring and measuring bias is still possible when researchers rely on heuristics and “gut-feelings” to analyze data.
- The possibility of using images or charts to inform natural language generation, as well as generating images and charts alongside text when storytelling using data.
- Ways that Chaturvedi and her team are trying to apply national language processing to clinical notes in electronic medical records, to potentially capture diagnoses or additional patient information.
- Augmented reality (AR) and virtual reality (VR) visualization projects that Sfazir and her team are researching in their lab, such as:
- Experimental studies that aim to understand how design guidelines for computers differ upon the implementation of AR or VR.
- The role of context in interpreting data.
- A way for researchers to project data as it is collected onto a physical field site to support data analysis in the environmental sciences.
- The use of AR for supporting analysis in emergency response, clinical therapy applications.
- The intrinsic, cognitive processes that permit humans to quickly catch visual anomalies that are not easily identified by machine.
- Experiences and challenges with privacy and accessibility when working with health data and visualization.
During the event, several attendees shared resources relevant to the conversation in the chat. A full list of educational programs, funding opportunities, on-campus support services, tools and software, and more shared during Carolina Data Science Now seminars can be found on our Resources page.
Our next seminar will be held on July 28 at 12 PM (ET). Register here.
Check out a playlist of our previous events here.