The rise of Big Data created the need for data applications to be able to consume data residing in disparate databases of wildly differing schema. The traditional approach to performing analytics on this sort of data has been to warehouse it; to move all the data into one place under a common schema so it can be analyzed. This approach is no longer feasible with the volume of data being produced, the variety of data requiring specific optimized schemas, and the velocity of the creation of new data. A much more promising approach has been based on semantic link data, which models data as a graph (a network of nodes and edges) instead of as a series of relational tables. To augment that approach, we've found that we can use machine learning to improve the semantic data models as the dataset evolves.

To augment that approach, we've found that we can use machine learning to improve the semantic data models as the data set evolves. Our specific use-case is text data in millions of documents. We've found that machine learning facilitates the storage and exploration of data that would otherwise be too vast to support valuable insights. Machine Learning (ML) allows for a model to improve over time given new training data, without requiring more human effort. For example, a common text-classification benchmark task is to train a model on messages for multiple discussion board threads and then later use it to predict what the topic of discussion was (space, computers, religion, etc).