7:00 – 4:00 VTX
- Continuing The Law of Group Polarization – done!
- Group polarization: A critical review and meta-analysis. Looks like a more rigorous version of TLoGP. It’s available in the library as a PDF if needed.
- Page 194: In short, the external materials and expert panels shift the argument pool available to the deliberators and are also likely to have effects on social influence.
- The way I read this, external trusted sources can shift the poles if they are incorporated into the discussion. Think about how a GPS affects wayfinding arguments. If search interfaces are modified such that they show the range of opinion and the position of the ‘Ten Blue Links’ within that range then, given its high system trust, we might expect individuals to adjust their belief trajectories based on their understanding of the pole’s position given the larger information landscape.
- Page 195: There are large lessons here about appropriate institutional design for
deliberating bodies. Group polarization can be heightened, diminished, and
possibly even eliminated with seemingly small alterations in institutional
- Starting The spreading of misinformation online
- Before continuing on backlinks, I spent some tom,e looking at the Microsoft Oxford system. LUIS is interesting, though I’m not sure exactly how to take advantage of it yet. I think this can be a chatbot construction kit? The WebLM system looks more immediately useful, kinda like AlchemyNLP. Maybe cheaper? You need a key, which you get here. And this is different from the Academic Knowledge API, which is also an Oxford project, but not listed on the Oxford site.
- Got the SrBacklinkObject persisting
- Adding backlinks to the ResultItemObject2 class. Whoops! Forgot that you have to set both relationships in a many-to-one:
- needed to split off the protocol from the curResult.link and add it back to the curResult.displayLink to get backlinks
- Done and working. Kinda like the fallback strategy.