We describe a machine-learning-based approach for extracting attribute labels from Web form interfaces. Having these labels is a requirement for several techniques that attempt to ...
We present a discriminative, latent variable approach to syntactic parsing in which rules exist at multiple scales of refinement. The model is formally a latent variable CRF gramm...
The provision of embedding neural networks into software applications can enable variety of Artificial Intelligence systems for individual users as well as organizations. Previous...
Information integration and retrieval are useful tasks in many information systems. In these systems, it is far from an easy task to directly integrate information from natural lan...
Abstract - We discuss an ensemble-of-classifiers based algorithm for the missing feature problem. The proposed approach is inspired in part by the random subspace method, and in pa...