Stardog brings machine learning to your data instead of forcing you to bring data to your machine learning. Stardog’s embedded machine learning allows data scientists to access the full breadth of unified data — structured, semi-structured, and unstructured.
Companies racing to simultaneously define and implement machine learning are finding, to their surprise, that implementing the algorithms used to make machines intelligent about a data set or problem is the easy part…. Instead, data is emerging as the key differentiator in the machine learning race.
- The Machine Learning Race Is Really a Data Race” MIT Sloan Management Review, December 14, 2018
Thanks to virtualization, training against the raw data is a cost-effective and scalable solution for improved model quality. when the source data changes, the built-in machine learning can quickly retrain and redeploy the models. Lastly, there is no risk to changing the underlying data, as Stardog protects from downstream writes.
Machine learning serves as a complement to the Inference Engine’s logical reasoning, which together provide a suite of reasoning capabilities that expose the full value of your connected data. Inference expresses all the implied relationships and connections between your data sources, creating a richer, more accurate view of your data. Better data means better machine learning. Additionally, constraints ensure data scientists only work with accurate, valid data. Use constraints to prevent the Knowledge Graph from accessing bad data or to simply flag inconsistencies in the data.
Stardog’s built-in machine learning offers both similarity search and predictive analytics (classification, regression). Similarity search is an important functionality when coupled with the connected data of a Knowledge Graph, as it can be used for pattern detection and recommendations. To learn how to use similarity search with Stardog, follow our tutorial.
How can you put machine learning to use with your Knowledge Graph?
- Enhance your semantic search application with predictive modeling to serve recommendations to your customers
- For drug discovery solutions, use machine learning to leverage the knowledge graph’s inherent traceability to predict pathways
- Monitoring for security incidents in your IT network? Use similarity search to figure out what incidents are most similar, and then do classification on those similar incidents
Furthermore, you can also use our python wrapper library, pystardog, to gain direct access to your Enterprise Knowledge Graph.
Learn how GeoPhy unified their data pipeline of 10,000 sources with Stardog, in order to automate commercial property valuations using ML. GeoPhy realized they needed a graph solution to enable massive integration and ingestion for this incredibly messy data.