Third @ieeevis workshop on TRust, EXpertise, and human factors in ML/AI-powered #Visualization systems. October 2022. #FlexAtTREX #TREXVIS2022 #ieeevis
- Tweets92
- Following11
- Followers81
- Likes21
ALT Headshot portrait of Dr. Simone Stumpf (University of Glasgow)
ALT We propose a three-layer provenance covering the user interface; the data used by the user interface; and the data sources from which the datasets used by the user interface were extracted. Tracking the data provenance, the user can identify issues in the visualization and in the data, download the data at each transformation step, and can identify the agent responsible for each activity (data transformation). Adding a communication support (e-mail or chat, for example) in the provenance visualization, the user can send messages to the agent responsible for each activity to ask questions, to make suggestions or to point errors.
ALT Two scatterplots depicting the same dataset. The left scatter plot is titled “High Perceptual Fluency.” The right scatter plot is titled “Low Perceptual Fluency.” Both scatter plots have tick marks on the x and y axes from 0 to 100 in steps of 10 (i.e. 0, 10, 20, etc). The tick marks on the x axes are labeled with dollar signs (i.e. $0, $10, $20, etc). The left scatter plot has light, grey, semi-transparent gridlines. The right scatter plot has dark, black, opaque gridlines. We hypothesize that the left scatter plot will be more highly trusted because it has higher fluency, thus potentially making perceptual fluency a proxy metric of trust
ALT T-rex 2022 workshop logo on the top left. On the top-right, the text reads: join us for a half day hybrid workshop to kick off IEEE Vis 2022. Find the program at: https://trexvis.github.io/Workshop2022/ TREX VIS 2022: The third workshop on TRust and EXpertise in Visualization. On the bottom left, the text reads: Keynote: When does the data “speak for itself”? Toward formal models of persuasive data visualization. Dr. Doug Markant, UNC-Charlotte. A picture of dr. Markant is also seen on the leftside of the text. On the bottom left, the text reads: Omni Hotel Oklahoma City Room: Oklahoma Station 8 Sunday 9am to 12pm Central The vis 2022 logo is seen on the bottom right corner of the image.
ALT T-rex 2022 workshop logo on the top left. On the top-right, the text reads: join us for a half day hybrid workshop to kick off IEEE Vis 2022. Find the program at: https://trexvis.github.io/Workshop2022/ TREX VIS 2022: The third workshop on TRust and EXpertise in Visualization. On the bottom left, the text reads: Keynote: When does the data “speak for itself”? Toward formal models of persuasive data visualization. Dr. Doug Markant, UNC-Charlotte. A picture of dr. Markant is also seen on the leftside of the text. On the bottom left, the text reads: Omni Hotel Oklahoma City Room: Oklahoma Station 8 Sunday 9am to 12pm Central The vis 2022 logo is seen on the bottom right corner of the image.
ALT This image represents a robot holding a hamburger.
ALT The simplified pipeline of FairAlign, a visual analytics platform for contextualized fairness assessment. A workflow example would be as follows: The laypeople sign up, choose an annotation dashboard and start taking decisions regarding algorithmic fairness, based on the provided visualizations of data and model's predictions. After the annotation process is finalized, data scientists and machine learning experts can login and analyze the values of the predefined fairness metrics, along with the aggregated results obtained from the human evaluation.
