Tag Archives: models

New York Times: Sentiment Analysis and Selling You Stuff

Something related to textual analysis:

The New York Times is researching how to contextualize which ads they show with the feelings an article is likely to inspire. I’m not a fan. They claim having learned that ads perform better on emotional articles won’t influence the newsroom, but we’ll see. At least they’re being transparent about doing this work. They’ve published an article with information on how they developed their sentiment analysis algorithm (link below).

There’s an explanation of the types of models they used and why. The initial steps were linear and tree-based textual analysis models, followed by a deep learning phase intended to “focus on language patterns that signaled emotions, not topics.” This outperformed the linear models some of the time, but not all of the time.

From what I can tell, the training set used a survey showing articles with images to establish a baseline, but the linear predictive models focus purely on text. I may be misunderstanding this or information may be missing. I expect that image selection can enhance or diminish the emotionality of an article. Perhaps sensational or graphic images would prove to drive more (or fewer) ad clicks. Despite the buffer the NYT cites between their newsroom and marketing arms, this feels like morally hazardous territory. So to answer the question in the title of the NYT piece, this article makes me feel disturbed. But I still didn’t click an ad.

It’s a quick read. Check it out.

https://open.nytimes.com/how-does-this-article-make-you-feel-4684e5e9c47

What the hell, Drucker?

This is a very hot take and I’m only up to page 13, but I’m posting this anyway.

I want to like this reading selection because I think it’s important to question how our cultural beliefs about logic and computing affect social structures… but I don’t like it. I think the introduction comes off as overwrought, self-serving hand-wringing and it’s really putting me off.

As I was reading this I thought of Richard Jean So’s article “All Models are Wrong.” The picture Drucker paints of the DH world is a model. Maybe in 2009 DH was as unthinking as she portrays, or maybe she distant read digital humanities projects without close reading the thinking around them to test her assumptions. I’m not in a position to say, with my 2018 perspective and only 5 weeks of studying the field. What I can point out is that this piece is riddled with absolutisms and sweeping declarations that strike me as iffy. To me, it feels like Drucker’s read of the DH field and DH projects lacks the very nuance, sensitivity and interpretism (whatever spellcheck, it’s a word if I want it to be) that she claims are missing in the DH work she critiques.

Drucker claims that consideration of design as a means of communication and usability, “plagues the digital humanities community” (p. 6). This is a cheap shot on my part, but has she actually used many DH tools? The user experience for many of them quite closely aligns with design as meditation, freestyle, or opportunity for idiosyncratic thinking.

Also what the hell is that weird conversation on page 12 where Drucker is trying to demonstrate that XML doesn’t communicate flirting?

In the example Drucker gives, a woman is bewildered and a man is “graciously” giving that woman knowledge and validation. The woman has big blue eyes that drop submissively as she blushes and asks him to guide her. No aspect of the man’s physical appearance is described. As always, it’s women who are fair game objects of a one-directional, sexualizing gaze.

I’m going to go stress eat about sexism (haha, I’m such a woman!). If, when I come back to this, further reading makes me reconsider these POVs, I’ll mea culpa in the comments.