Implicit bias, rooted in unconscious attitudes, fuels discrimination based on race, gender, or sexual orientation, disproportionately impacting marginalized groups. Despite the development of interventions addressing provider awareness of implicit bias, the advancement of clinical education through technology has been slow. In the UnBIASED research project, we investigated the usability of ConverSense, a personalized communication assessment tool to raise healthcare providers' awareness of bias in their communication with patients. This web-based tool measures social dimensions such as warmth, interactivity, engagement, and assertiveness from recorded patient-provider visits, and visualizes these patterns through graphs and embedded clips. In this study, we (PT, NE) examined whether ConverSense meets usability standards through heuristic evaluations conducted by design experts. Six healthcare technology experts participated in the evaluation of ConverSense using Nielsen's 10 usability heuristics. Experts documented usability issues for each heuristic and rated their severity on a scale from 0 (not a problem) to 4 (catastrophic problem). Through our analysis, we (PT, NE) identified three cross-cutting themes: 1) Poor design, where experts noted the absence of undo or delete buttons, making navigation challenging, and the distracting color scheme on graphs; 2) Data visualization issues, with experts expressing difficulty interpreting charts and uncertainty about what is considered ideal or good communication. One expert said “It's unclear what is considered ideal/good…for each gauge chart, high interactivity, engagement, and warmth I would assume are ideal/good. But how the charts are displayed in the system I cannot know for certain”; 3) Ambiguity in information presentation, where experts sought more definitions for measured social dimensions and recommended training links to help them identify personal actions they can take to improve their communication. This study underscores the value of incorporating expert feedback and addressing usability issues to improve tools like ConverSense to address implicit bias and promote equitable patient-provider interactions.

Geoparsing, the task of assigning coordinates to locations extracted from text, enables us to better understand how places change over time through historical documents. The task has seen many advancements with the growth of neural machine learning methods, but these models require large amounts of training data. In the history domain, many geoparsing corpora are from newspaper collections because of the availability of labeled data. However, not all historical research benefits from rich data availability, so we seek to understand and improve existing geoparsing methods for small corpora. In this project, I focus on the Svoboda Diaries, a collection of personal diaries written by British steamship purser Joseph Svoboda during the late 19th century in Ottoman Iraq. How do geoparsing methods perform in the Svoboda Diaries, a small historical text corpus written primarily in English, but also including text in Arabic, Italian, and French? We develop a map-based generate-and-rank approach with clustering of context words surrounding each location in the text. The location data is retrieved from two knowledge bases: GeoNames, a gazetteer with over 25 million geographical names, and Wikidata, a free and collaboratively edited database. We first extract the locations from the text using a natural language processing method called named entity recognition. Since modern location names may differ from historical spellings, we then generate alternate names for each identified location through translation and hand-crafted Romanization rules, such as “Basra” for “Basreh”, and query knowledge bases to retrieve coordinate data associated with these names. We find that our method to augment transliterated Arabic names with alternate names helps with gazetteer lookups, suggesting potential for augmenting data in other small corpora tasks. Our research with the Svoboda Diaries expands on existing geoparsing methods and can apply to other digital humanities projects facing the small corpora challenge.