A neural network can identify references that are unlikely to support an article’s claims, and scour the web for better sources.

Should I believe this headline?

@salarua@sopuli.xyz
link
fedilink
99
edit-2
10M

Wikipedian here - AI on Wikipedia is actually nothing new. we’ve had a machine learning model identify malicious edits since 2017, and Cluebot (an ML-powered anti-vandalism bot) has been around for even longer than that.

even so, this is pretty exciting. from what i gather, this is a transformer model turned on its side; instead of taking textual data and transforming it, it checks to see if two pieces of textual data could reasonably be transformations of each other. used responsibly, this could really help knock out those [dubious] and [failed verification] tags en masse

If I’m understanding you correctly, it doesn’t ever edit the actual pages, it just adds flags on certain kinds of content. Is that right?

yes. it only surfaces citations that may back up the content better, an editor still has to read the source and approve the change

Fascinating, as a developer, where can I read more/contribute?

@HalJor@beehaw.org
link
fedilink
English
81Y

The aforementioned ClueBot is here: https://en.wikipedia.org/wiki/User:ClueBot_NG

For bots in general, start here: https://en.wikipedia.org/wiki/Wikipedia:Bots

Create a post

A nice place to discuss rumors, happenings, innovations, and challenges in the technology sphere. We also welcome discussions on the intersections of technology and society. If it’s technological news or discussion of technology, it probably belongs here.

Remember the overriding ethos on Beehaw: Be(e) Nice. Each user you encounter here is a person, and should be treated with kindness (even if they’re wrong, or use a Linux distro you don’t like). Personal attacks will not be tolerated.

Subcommunities on Beehaw:


This community’s icon was made by Aaron Schneider, under the CC-BY-NC-SA 4.0 license.

  • 1 user online
  • 56 users / day
  • 167 users / week
  • 618 users / month
  • 2.31K users / 6 months
  • 1 subscriber
  • 3.28K Posts
  • 67K Comments
  • Modlog