#28: MLDublin meets @ DogPatch Labs- 1 min
A review of what has been done in the past couple of year to get the best performance from neural lanuage models. From embeddings wieghts, hyperparameter search, to recurrent residual connections.
Entity relation extraction is a problem in a similar area to entity link and entity disambiguation and is fundamental task in information extraction. Entity relation extraction is a useful for extraction of structured data from unstructured data, like raw text.
We envision a future where journalists will no longer be limited to report pastor current affairs, but they will be empowered by Artificial Intelligence and Machine Learning (AI/ML)to write about future events with a fair degree of certainty.In this talk we present "Minerva", a tool to predict and visualize the (non-obvious) interconnections of global risks that will be at the core of tomorrow's news. We describe how Minerva leverages news data collections available in the Web and applies AI/ML to discover the multiple-relations among global risks. Minerva's data-driven approach is more appealing, in terms of timeliness and efficient discovery of global risks relations, than current annual reports based on opinion surveys offered by the World Economic Forum.