Fine-Grained Entity Type Classification by Jointly Learning Representations and Label Embeddings

Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 797–807, Valencia, Spain, April 3-7, 2017. c⃝2017 Association for Computational Linguistics Fine-Grained Entity Type Classification by Jointly Learning Representations and Label Embeddings Abhishek, Ashish Anand and Amit Awekar Department of Computer Science and Engineering Indian Institute of Technology Guwahati Assam, India - 781039 {abhishek.abhishek, anand.ashish, awekar}@iitg.ernet.in Abstract Fine-grained entity type classification (FETC) is the task of classifying an entity mention to a broad set of types. Distant su- pervision paradigm is extensively used to generate training data for this task. How- ever, generated training data assigns same set of labels to every mention of an en- tity without considering its local context. Existing FETC systems have two major drawbacks: assuming training data to be noise free and use of hand crafted fea- tures. backs.Our work overcomes both draw- We propose a neural network model that jointly learns entity mentions and their context representation to elim- inate use of hand crafted features. Our model treats training data as noisy and uses non-parametric variant of hinge loss function. Experiments show that the pro- posed model outperforms previous state- of-the-art methods on two publicly avail- able datasets, namely FIGER(GOLD) and BBN with an average relative improvement of 2.69% in micro-F1 score. Knowledge learnt by our model on one dataset can be transferred to other datasets while us- ing same model or other FETC systems. These approaches of transferring knowl- edge further improve the performance of respective models. 1 Introduction Entity type classification is the task for assigning types or labels such as organization, location to entity mentions in a document. This classifica- tion is useful for many natural language process- ing (NLP) tasks such as relation extraction (Mintz et al., 2009), machine translation (Koehn et al., 2007), question answering (Lin et al., 2012) and knowledge base construction (Dong et al., 2014). There has been considerable amount of work on Named Entity Recognition (NER) (Collins and Singer, 1999; Tjong Kim Sang and De Meul- der, 2003; Ratinov and Roth, 2009; Manning et al., 2014), which classifies entity mentions into a small set of mutually exclusive types, such as Person, Location, Organization and Misc. How- ever, these types are not enough for some NLP applications such as relation extraction, knowl- edge base construction (KBC) and question an- swering. In relation extraction and KBC, know- ing fine-grained types for entities can significantly increase the performance of the relation extrac- tor (Ling and Weld, 2012; Koch et al., 2014; Mitchell et al., 2015) since this helps in filtering out candidate relation types that do not follow the type constrain. Fine-grained entity types provide additional information while matching questions to its potential answers and significantly improves performance (Dong et al., 2015). For example, Li and Roth (2002) rank questions based on their ex- pected answer types (will the answer be food, ve- hicle or disease). Typically, FETC systems use over hundred la- bels, arranged in a hierarchical structure. …

Content truncated to 500 words