Shikhar Vashishth

We propose RESIDE, a distantly-supervised neural relation extraction method which utilizes Graph Convolution Networks (GCN) to encode syntactic information from text for improved relation extraction. Expand

We propose Canonicalization using Embeddings and Side Information (CESI) -- a novel approach which performs canonicalization over learned embeddings of Open KBs. Expand

We perform an extensive reexamination study of recent neural network based KGC techniques and find that many such models have issues with their score functions. Expand

We propose CompGCN, a novel Graph Convolutional framework for multi-relational graphs which jointly embeds both nodes and relations in a relational graph. Expand

We propose SynGCN, a flexible Graph Convolution based method for learning word embeddings, which outperforms existing methods on various intrinsic and extrinsic tasks. Expand

The attention layer in a neural network model provides insights into the model's reasoning behind its prediction, which are usually criticized for being opaque. Expand

We propose NeuralDater, a Graph Convolutional Network (GCN) based document dating approach which jointly exploits syntactic and temporal graph structures of document in a principled way. Expand

We propose InteractE, a model that applies convolutional filters on 2D reshapings of entity and relation embeddings in order to capture rich interactions between them. Expand

We propose ConfGCN, a GCN framework for graph-based SSL, which estimates labels scores along with their confidences jointly in GCN-based setting. Expand

We present an approach for monitoring and measuring online per-flow and per-port packet loss and its comparison with some of the novel methods for measuring packet loss proposed for SDN and OpenFlow networks. Expand