Topicalizer: reframing core concepts in machine learning visualization by co-designing for interpretivist scholarship

Click to increase image sizeClick to decrease image size

Acknowledgments

Thanks to the additional students who made various contributions to the coding for this visualization along the way: Yin Luo, Hannah Lambert, Alex Van Heest, and Rachel McCoog; to David Mimno for helpful conversations; and to the bloggers for sharing their stories.

Notes

¹ The astute reader may notice a similarity between visualization design studies and research through design (Zimmerman and Forlizzi 2014), but a full comparison of these approaches exceeds the scope of this paper.

² https://mohistory.org/collections/item/resource:103591 and https://bit.ly/2XeDid5.

³ Two other visualizations, TOME (Klein et al., 2015) and Serendip (Alexander et al., 2014), also seemed attractive, but at the time of commencing the work presented here their source code was not publicly available.

⁴ This paper uses “quotes” for the names of topics, derived from the topic’s high probability words, and Small Caps for researchers’ manually-assigned topic labels.

⁵ At the time this work was conducted, techniques for biasing topic models away from such known structure (Thompson & Mimno, 2018) had not yet been published.

⁶ Upon publication, the current version of the source code for Topicalizer will be made publicly available at https://github.com/ericpsb/topicalizer. This version will include data to generate and run the Gunn corpus visualization. However, it will exclude the ASD corpus due to concerns around the bloggers’ identities and the sensitive nature of such data (cf. Baumer & McGee, 2019).

Additional information

Funding

This material is based on work supported in part by the NSF under Grant No. IIS-1844901.

Notes on contributors

Eric P. S. Baumer

Eric P. S. Baumer ([email protected], ericbaumer.com) studies human interactions with NLP and AI algorithms in the context of social computing systems; he is an Assistant Professor of Computer Science & Engineering at Lehigh University.

Drew Siedel

Drew Siedel is an alumnus from Computer Science & Engineering at Lehigh University.

Lena McDonnell

Lena McDonnell is an alumna from Computer Science & Engineering at Lehigh University.

Jiayun Zhong

Jiayun Zhong is an alumna from Computer Science & Engineering at Lehigh University.

Patricia Sittikul

Patricia Sittikul is an alumna from Computer Science & Engineering at Lehigh University.

Micki McGee

Micki McGee ([email protected], https://www.fordham.edu/info/20855/faculty/5023/micki_mcgee) is a sociologist and cultural critic interested in disability studies and digital humanities research methods; she is an Associate Professor of Sociology at Fordham University.