New statistical tool MINE (Maximal Information-based Nonparametric Exploration) developed by two brothers from Harvard and MIT to allow searching for relationships in large datasets. http://www.exploredata.net/
I am very excited. I’ve been waiting for something like this for years. There are now three things that I would like to happen:
1. All ecological data from peer-reviewed journals should be pooled into a giant dataset, so that previous research can be used to find correlations.
2. Public contribution to the dataset (e.g. soil pH, species surveys, etc.). In order for this to work, I think data entrants would have to be ranked so that you can search for peer-reviewed data and/or public data, and so that data has varying clout depending on the source type.
3. MODELS. SO MANY MODELS. Automatically computed based on data and data relationships over time and updated whenever someone adds to the dataset.