3 ms·
It doesn't seem like citations would be good for discovery, because there must be a significant latency between when a paper is released and when citations star
by anamexis 2y ago
It doesn't seem like citations would be good for discovery, because there must be a significant latency between when a paper is released and when citations start coming in.
- bee_rider 2y agoProbably it would be best to just get a site on the web and expose a bunch of different metrics so people can sort by whatever. Citations are probably not the best metric for discovery, but also this really just makes me wonder if papers are not the best thing for discovery. An academic produces ideas, not papers, those are just a side-effect. The path is something like: * make a idea * write short conference papers about it * present it in conferences * write journal papers about it * maybe somebody writes a thesis about it (Talking to people about it throughout). If we want to discover ideas as they are being worked on, I guess we’d want some proxy that captures whether all that stuff is progressing, and if anybody has noticed… Finding that proxy seems incredibly difficult, maybe impossible.
- michaelmior 2y agoI'm not sure I agree about papers just being a side effect. An idea by itself has significantly less value than an idea which has been clearly documented and evaluated. I think a paper is often still the best way to do this.