Multi Label Semantic Relation Classification

1378 Words Sep 24th, 2016 6 Pages
Multi-label Semantic Relation Classification
Between Pair of Nominals
Kartik Dhiwar,
PG Scholar, Department of Computer Science and Engineering, SSGI, SSTC, Bhilai (CG), India Abhishek Kumar Dewangan
Professor, Department of Computer Science and Engineering, SSGI, SSTC, Bhilai (CG), India Abstract: Relation classification is a keynote in the field of Natu-ral Language Processing (NLP) to mine information from text facing problems of over-reliance on the standard of handcrafted features. Features annotated by specialists and lin-guistic data derived from linguistic analysis modules is expen-sive and ends up with the difficulty of error propagation. Rela-tion extraction plays a crucial role in extracting struc-tured data from unstructured sources like raw text. One might want to seek out interactions between medicines to create medical information or extract relationships among people to create a simply searchable knowledgebase. We propose a deep Convolutional Neural Network model for the multi-label text relation classification task without hand crafted features. This model outperforms the best existing model as per our knowledge without depending much on manually engineered features with the small updates in the loss function applied.
Index Terms – Relation Classification, Features, Label, Convolutional Neural Network, Information Extraction.
Natural Language Processing tasks are now applicable to…
Open Document