(A thought experiment of sorts, the inspiration which is:
https://www.bbc.com/news/world-46453084)<p>I was wondering if anyone was aware of projects to enable "crowd-solving" of crimes (starting with cold-cases or such), or has thought about how this could be achieved (e.g. the kind of platform required, legal/privacy issues, etc.)?<p>By 'platform' I mean a system that would complement a true-crime podcast (in the above link) and represent the constituent parts or tasks of a case (based on 'crime-scene'- and 'legal' -logic), which can push tasks -- mechanical-turk-ing them -- onto existing community spaces before aggregating results:<p>-Some tasks might just require OCR'ing of documents (or parts-thereof) and can be farmed-off to a Recaptcha system. They could similarly require the making of small/simple cartoony animations (or 'model making') to illustrate how a crime was done for example. These as tasks to do with preparing/modelling the evidence<p>-Other tasks might require some data analysis (e.g. financial data), can can be made into a Kaggle or 'Data Science for Social Good' project.<p>-Other tasks might be fact finding tasks that require a basic google/wolfram-alpha search or something more and might just require short answers; e.g. "how long is the walk from the suspect's location to the victim's house"<p>-Other tasks/puzzles/problems will require more thought or time and can perhaps be presented through stackexchange or reddit style communities (and not just formats) to facilitate discussion; e. g. "why would someone walking on street x in time y take 3 minutes longer than others do?"<p>And we'd need the platform to keep track of these discussions and combine the results over time to deduce the overall status.<p>Thoughts?<p>[Of course this could all be seen as the preliminary stage for a 'Minority Report' type of world, but relying more-so on human computation and focusing initially on past crimes)