5. Classifying Meanings & Documents - [E1] 1. Cheng, Justin, Michael Bernstein, Cristian Danescu-Niculescu-Mizil and Jure Leskovec. 2017.

[JunsolKim]

Post questions here for this week's exemplary readings: 1. Cheng, Justin, Michael Bernstein, Cristian Danescu-Niculescu-Mizil and Jure Leskovec. 2017. “Anyone Can Become a Troll: Causes of Trolling Behavior in Online Discussions.” WWW 2017: 1-14.

[konratp]

I find it interesting that the authors opted for an approach that supplements results from an experiment with inferences made from through large scale text analysis. How commonly used is large-scale textual analysis as a supplementary, rather than primary method? I also wonder what exactly the benefits are of conducting an experiment instead of studying a greater variety of samples, not just one website (CNN)? Especially given the limitations the authors address in the study, I have doubts about the accuracy of the results from the experiment.

[Jiayu-Kang]

Interesting read! I like how the authors discussed how mood and surrounding context affect trolling behavior online. I do wonder to what extent their findings can be generalized to our daily life: this study seems to answer the question "when/why would people engage in trolling behavior when they have to make a comment", while in reality, what we care seem to be "who would engage in online trolling behavior and why." Can we say that, the fact that the participants are asked to comment (versus in reality they choose to comment or not) is also a special condition that contributes to trolling behavior that those participants might not normally engage in?

[Qiuyu-Li]

I feel that this is not a novel idea, but it's still cool how they detect and "watch a person get annoyed and become a troll." Here are just some points I hope further researchers might address, and they may be helpful for us to understand more of online trolling behaviors, as well as reflect on ourselves:

• How trolling one can be? In particular, is there any difference between those who have been used to trolling language, and those who are only ignited in a particular situation?
• How long does the trolling behavior last? For example, is it like some kind of epidemic, that when a person is ignited at some time, she will continue to act troll? Or do those who continue to act troll have some other inner characteristics?

[sudhamshow]

I was wondering about the associated Hypothesis - if trolling can so easily predicted and mitigated by the type of circumstance one is involved in. One can also troll playfully - for e.g., many rhetorical questions on reddit and twitter are met with trolls which are done in good faith. The moderators are not obliged to take any actions on these as well. How do the authors explain this behaviour?