You are viewing a single comment's thread from:
RE: SteemOps: Extracting and Analyzing Key Operations in Steemit Blockchain-based Social Media Platform - ACM
You can download the PDF for free from arXiv, here - https://arxiv.org/pdf/2102.00177.pdf.
You can download the PDF for free from arXiv, here - https://arxiv.org/pdf/2102.00177.pdf.
Thanks. I have downloaded it. I am happy there are only 6 pages. Research papers are notoriously known for their copious length. With 6 pages I will be done in no time should I start going through them.
I agree about the page length. This one is actually a pretty quick read. The previous one that I linked to has something like 22 pages, and I still haven't finished it.
I can understand...With those small font sizes and the 22 pages, reading it may even feel like a punishment. Under such situations, note taking may be required to keep track of the bigger picture of the publication.
I finally made it through, but it took a while. It was an interesting article. My favorite idea was an idea to add a capability for users to (positively) flag comments as "constructive" in order to raise the constructive commentary above the trolls.
One thing that is really bizarre to me is that survey respondents seemed quite comfortable with submitting to censorship.
That's too big of an ask for a typical user, though. Automating it would likely not bee too difficult, and would get the job done without depending on users to do the work.
Probably true, but I think that automating it might also turn out to be pretty challenging. AI isn't great at dealing with nuance yet.
Maybe a hybrid model, where you put the option there but recognize that some people will be better at it (and more willing to participate) than others, so you augment user input with AI.
Hybrid sounds right. A human (or humans?) could also validate or invalidate algorithmic evaluations since the algorithms might miss sarcasm, context clues etc. If it's incentivized, it might reduce the hurdle of the high ask.
Still, I think natural language processing is far enough along, and there are open source projects that a dev could copy to automate something like this more easily than you suggest. Imperfect example, but: there have to be bunches of projects like this demo Tone Analyzer that uses Watson in open source repositories...
Something like this demo might not demonstrate completely accurate results, but wouldn't custom training input allow somebody to make a better model?
This is an excellent idea. Quite often, people use the words troll and constructive criticism interchangeably, but they are two different things. These labels will help them stand apart.
Also, censorship on a decentralized network seems out of place to me.