About this event
New communities are growing on the internet, but they are using decentralized architectures, peer-to-peer protocols, and cultural codes (emojis, meme gifs, etc) that make it harder to survey, map, and understand. Your challenge is to explore this new ecosystem and develop novel data science techniques to illuminate it.
You will be provided with a list of platform and protocol sources (Check the Rules for a comprehensive set of links to the data and other resources provided for this challenge.)
You must:
Scrape and ingest a representative set of content from each resource provided.
Perform text analysis (including entity extraction/identification, sentiment, etc) on all unstructured and structured text fields. This includes translating from the source language to English (at some point in the overall analytical pipeline)
Perform graph/SNA-type analysis on all relationship data (post, author, comment, etc)
Perform analysis on all media posted. Goal is to derive as much context as possible from a given media post as well as context of the collective post (media, non-media such as tags, comments, etc), and any other topical/contextual "channels".
Finally, you will have to provide an overall "contextual dashboard" to illustrate the topical/semantic/sentimental trends within and across different platforms and protocols. Extra credit for a "Grey Network Weather Report" that can illustrate probabilistic correlation over time series across all sources ingested.
This challenge is sponsored by Donovan Group, SOCOM J-5, and will be judged by Maj Jen Snow and CDR Ed "Rockhound" Ward.
Schedule
- Friday
-
Schedule for the day is not published
- Saturday
-
Schedule for the day is not published
Rules
"Rules are for people without principles" - M. Scott Here are the OpenWERX Principles: Open Source We recommend that all your work for this Hackathon be made publicly available under an Open ...