Hope You Are Well is a series of visualizations that analyze the language of my sent e-mails over time. Together, they reveal a surprisingly personal narrative. Data analysis becomes a tool for introspection.
Background
Technical:
LIWC: Linguistic Inquiry and Word Count (liwc.net); Google BigQuery and Refine; Machine Learning for Hackers by John Myles White
Projects/Art: "ReConstitution 2012;" Steven Wolfram's "The Personal Analytics of My Life;""Moveable Type" by Mark Hansen and Ben Rubin; "Language Is Not Transparent" by Mel Bochner
Books: The Test Drive by Avital Ronell; I Swear I Saw This by Michael Taussig; Mindset by Carol Dweck
Audience
This is a personal project. Its universal appeal is that everyone uses e-mail.
User Scenario
Not applicable; best viewed in a gallery setting or online.
Implementation
My project harnesses my e-mail archive, reconstructing past experience. After parsing all my sent messages since 2005, I found that my work and personal e-mails followed particular structures. During crisis, their distinct qualities started to blur.
"Hope You Are Well" is an expression of my research process. In order to understand the patterns I was seeing, I created diagrams that categorize my phrasing. I then supplemented my interpretations with linguistic analysis (primarily sentiment). I was interested in finding the gaps between computational analysis and my own biases. I also wanted to see if there was any computational basis for which e-mails I'm willing to forgive myself for, and which ones I'm still embarrassed by.
Conclusion
I didn't know that my e-mails followed any sort of pattern. After separating out the general formats for work and personal e-mail, I found that in moments of crisis, I go into great detail about why I'm writing and take on a defensive tone. I thought this sort of message was an aberration— that it only occurred in the past year or two. Through looking at seven years's worth of mail, I learned that I've been writing these same, anxious e-mails all along.
Background
Technical:
LIWC: Linguistic Inquiry and Word Count (liwc.net); Google BigQuery and Refine; Machine Learning for Hackers by John Myles White
Projects/Art: "ReConstitution 2012;" Steven Wolfram's "The Personal Analytics of My Life;""Moveable Type" by Mark Hansen and Ben Rubin; "Language Is Not Transparent" by Mel Bochner
Books: The Test Drive by Avital Ronell; I Swear I Saw This by Michael Taussig; Mindset by Carol Dweck
Audience
This is a personal project. Its universal appeal is that everyone uses e-mail.
User Scenario
Not applicable; best viewed in a gallery setting or online.
Implementation
My project harnesses my e-mail archive, reconstructing past experience. After parsing all my sent messages since 2005, I found that my work and personal e-mails followed particular structures. During crisis, their distinct qualities started to blur.
"Hope You Are Well" is an expression of my research process. In order to understand the patterns I was seeing, I created diagrams that categorize my phrasing. I then supplemented my interpretations with linguistic analysis (primarily sentiment). I was interested in finding the gaps between computational analysis and my own biases. I also wanted to see if there was any computational basis for which e-mails I'm willing to forgive myself for, and which ones I'm still embarrassed by.
Conclusion
I didn't know that my e-mails followed any sort of pattern. After separating out the general formats for work and personal e-mail, I found that in moments of crisis, I go into great detail about why I'm writing and take on a defensive tone. I thought this sort of message was an aberration— that it only occurred in the past year or two. Through looking at seven years's worth of mail, I learned that I've been writing these same, anxious e-mails all along.