Series
Career, Research, and Innovation Development Conference

Permanent Link
https://hdl.handle.net/1853/70851
Series Type
Event Series
Description
Associated Organization(s)
Associated Organization(s)
Organizational Unit

Filters

2021 - 2024 3
3
3
3
3
3
Reset filters

Settings
Author: Narechania, Arpit ×

Publication Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Increasing Awareness of Human Biases during Visual Data Analysis using Visual and Haptic Feedback
    (Georgia Institute of Technology, 2024-02-08) Narechania, Arpit ; Paden, Jamal ; Endert, Alex
    Human biases impact the way people analyze data and make decisions. Women denied C-suite promotions (gender bias), ailing but younger people denied optimal treatment (age bias), dark-skinned people denied parole (racial bias), etc. are examples of biases rampant in the world. Visual data analysis tools such as Tableau and Excel help users see and understand their data but do not report potential biases exhibited by users (e.g., an overemphasis on the Age attribute). Existing research tools have explored visual means (e.g., highlighting the Age attribute to appear darker than other attributes) to increase users' awareness about (biased) analytic behaviors. We believe that using a single, "visual" modality to present such information is a passive type of guidance that only burdens the user's perception skills, which are already engaged to perform the analysis task. We investigate how a more active type of guidance, a combination of "visual" and "haptic" modalities, can better guide the user. We present a visual data analysis system, wired to a haptic gaming mouse. This enhanced system tracks a user's interactions with data and presents them back via haptic feedback (e.g., "buzz"es or vibrates the mouse whenever bias is detected). Through an exploratory user study, we find that these dual guidance modalities can sometimes actively stimulate and engage the user's attention, making them more aware of their analytic behaviors. However, we also find that the haptic feedback can also distract the user, informing the design of future multimodal guidance-enriched user interfaces.
  • Thumbnail Image
    Item
    vitaLITy: Promoting Serendipitous Discovery of Academic Literature
    (Georgia Institute of Technology, 2022-01-27) Narechania, Arpit ; Karduni, Alireza ; Wesslen, Ryan ; Wall, Emily
    There are a few prominent practices for conducting academic literature reviews, including searching for specific keywords on Google Scholar or checking citations from initial seed paper(s). While these approaches serve a critical purpose for academic literature reviews, there remain challenges in identifying relevant literature when (1) different work may utilize the same terminology (e.g., “transformer” in electronics refers to a device that transfers energy between circuits; whereas in computing, it refers to a type of deep learning model, commonly applied to unstructured text data) or (2) similar work may utilize different terminology (e.g., work on “bias” in visualization seldom mentions “uncertainty” even though bias sometimes emerges when people make decisions under uncertainty). We developed a visual analytics system, VitaLITy, to promote serendipitous discovery of academic papers wherein users may “stumble upon” relevant literature, when other search approaches may fail. VitaLITy (1) utilizes transformer language models to help users find semantically similar papers given a list of seed paper(s) or a working abstract, (2) visualizes the embedding space in an interactive 2-D scatterplot, and (3) summarizes meta information about the paper corpus (e.g., keywords, co-authors, citation counts, and publication year). We also curated a comprehensive dataset comprising papers from 38 popular visualization publication venues (e.g., ACM CHI, IEEE VIS) using custom web-scrapers. We have open-sourced the VitaLITy system, dataset, and web-scrapers at https://vitality-vis.github.io/ for the research community to grow the list of supported venues, potentially expanding into other fields, e.g., biology.
  • Thumbnail Image
    Item
    Lumos: Increasing Awareness of Biases during Visual Data Analysis
    (Georgia Institute of Technology, 2021) Narechania, Arpit ; Coscia, Adam ; Wall, Emily ; Endert, Alex
    Human biases impact the way people analyze data and make decisions. Dark-skinned people denied parole (racial bias), women denied C-suite promotions (gender bias), ailing but younger people denied optimal treatment (age bias), etc. are examples of biases rampant in the world. Visual data analysis tools such as Tableau and Excel help users see and understand their data but do not report potential biases exhibited by users (e.g., an overemphasis on the Age attribute). Lumos is an analysis tool that helps users visualize traces of their interactions with data to increase awareness of potential biases. Using in-situ and ex-situ visualization techniques, Lumos provides real-time feedback to users to reflect upon their activities and potentially change future course. For example, Lumos remembers and highlights datapoints that have been previously examined in the same visualizations (in-situ) and overlays the interacted datapoints on the underlying data distribution in a separate visualization (ex-situ). Now sometimes, custom policies rather than biases drive decision-making. For example, a university admissions committee selecting more female than male student applicants can be a conscious choice in abidance to the university's gender-equality policy, rather than an unconscious bias. To address these situations, Lumos allows users to configure custom target distributions and accordingly updates the interaction traces. We believe Lumos can improve data exploration and decision-making scenarios to not only help mitigate the dangers of human biases affecting judgements, but also foster more transparent analysis processes.