Unsupervised Word and Dependency Path Embeddings for Aspect Term Extraction
This work addresses aspect term extraction for natural language processing applications, offering an incremental improvement through novel embedding features.
The paper tackled aspect term extraction by developing an unsupervised method that learns embeddings for words and dependency paths, optimizing w1 + r = w2 in a low-dimensional space using a recurrent neural network. Experimental results on SemEval datasets showed state-of-the-art performance with only embedding features, outperforming other methods by incorporating syntactic information.
In this paper, we develop a novel approach to aspect term extraction based on unsupervised learning of distributed representations of words and dependency paths. The basic idea is to connect two words (w1 and w2) with the dependency path (r) between them in the embedding space. Specifically, our method optimizes the objective w1 + r = w2 in the low-dimensional space, where the multi-hop dependency paths are treated as a sequence of grammatical relations and modeled by a recurrent neural network. Then, we design the embedding features that consider linear context and dependency context information, for the conditional random field (CRF) based aspect term extraction. Experimental results on the SemEval datasets show that, (1) with only embedding features, we can achieve state-of-the-art results; (2) our embedding method which incorporates the syntactic information among words yields better performance than other representative ones in aspect term extraction.