Neural Relation Extraction via Inner-Sentence Noise Reduction and Transfer Learning
This addresses noisy data issues in knowledge base completion for NLP researchers, but it is incremental as it builds on existing distant supervised methods.
The paper tackled the problem of low-quality sentences with noisy words in distant supervised relation extraction, proposing a word-level approach that improved the Precision/Recall area from 0.35 to 0.39 over state-of-the-art methods on NYT and Freebase datasets.
Extracting relations is critical for knowledge base completion and construction in which distant supervised methods are widely used to extract relational facts automatically with the existing knowledge bases. However, the automatically constructed datasets comprise amounts of low-quality sentences containing noisy words, which is neglected by current distant supervised methods resulting in unacceptable precisions. To mitigate this problem, we propose a novel word-level distant supervised approach for relation extraction. We first build Sub-Tree Parse(STP) to remove noisy words that are irrelevant to relations. Then we construct a neural network inputting the sub-tree while applying the entity-wise attention to identify the important semantic features of relational words in each instance. To make our model more robust against noisy words, we initialize our network with a priori knowledge learned from the relevant task of entity classification by transfer learning. We conduct extensive experiments using the corpora of New York Times(NYT) and Freebase. Experiments show that our approach is effective and improves the area of Precision/Recall(PR) from 0.35 to 0.39 over the state-of-the-art work.