[18:55:33] dsaez: (in case you're still around) have you read https://cyber.harvard.edu/publication/2018/network-propaganda ? [18:56:13] (it's a full book;) [18:56:13] leila, nop, I'll check [18:56:37] dsaez: check and if you find it helpful for your work, feel free to buy it. [18:56:47] (I see good reviews for it. will check it myself as well.) [19:54:31] isaacj: nice job with the presentation. [19:55:01] thanks! [19:55:35] isaacj: the main question for me is: what does it take to almost fully automatize the analysis. I /think/ the biggest steps that need human intervention are "quality of translations" and "debiasing", the latter partially being an art. but maybe we can overcome at least the latter? [19:58:11] leila: yeah, the pipeline has certainly gotten much smoother. it'll always take someone to run it though (there are always some manual tweaks that need to be accounted for, like in the case of our upsampling in French/English). it's possible though that we get to a point where we can say if you account for session length alone and make a standard correction, you're probably capturing 90% of what the full debiasing would [19:58:40] yup [20:34:50] leila, please check this paper: https://link.springer.com/epdf/10.1007/s42001-017-0005-6 [20:35:16] that might be a good journal to submit the work about missinformation [20:35:53] they don't list surveys as one of the topics, but might be a possition paper or something like that [20:56:08] dsaez: yup. had a look and it can be an option. As you said earlier, the choice of journal will depend on the few areas of focus and expansion of the paper as well, and what the strongest area of it will be. Just as an example, if a good chunk of the focus ends up being on policy, international affairs can be an option, while for more CS focused directions (current inclination), the journal you linked can work. [21:06:28] I don't think that I have the knowledge to write a paper in a good journal of international affairs. I'm a Computer scientist. [21:07:00] dsaez: it was an example. ;)