Four short links: 14 Feb 2017
Rapping Neural Network, H1B Research, Quantifying Controversy, Social Media Research Tools
- Rapping Neural Network — It’s a neural network that has been trained on rap songs, and can use any lyrics you feed it and write a new song (it now writes word by word as opposed to line by line) that rhymes and has a flow (to an extent). With examples.
- H1B Research — H1B holders are paid less and often weaker in skills compared to their American counterparts.
- Amazon Chime — interesting to see a business service from Amazon, not a operations service. This is better (they claim) meeting software: move between devices, with screen-sharing, video, chat, file-sharing.
- Quantifying Controversy in Social Media — The research is carried out in the context of Twitter, but in theory can be applied to any social graph structure. A topic is simply defined as a query, often a hashtag. Given a query, we can build a conversation graph with vertices representing users, and edges representing activity and interactions between users. Using a graph partitioning algorithm, we can then try to partition the graph in two. If the partitions separate cleanly, then we have a good indication that the topic is controversial and has polarized opinions.
- Social Media Research Toolkit — a list of 50+ social media research tools curated by researchers at the Social Media Lab at Ted Rogers School of Management, Ryerson University. The kit features tools that have been used in peer-reviewed academic studies. Many tools are free to use and require little or no programming. Some are simple data collectors such as tweepy, a Python library for collecting Tweets, and others are a bit more robust, such as Netlytic, a multi-platform (Twitter, Facebook, and Instagram) data collector and analyzer, developed by our lab. All of the tools are confirmed available and operational.
