Towards Robust Text Classification with Semantics-Aware Recurrent Neural Architecture

摘要

Deep neural networks are becoming ubiquitous in text mining and natural languageprocessing, but semantic resources, such as taxonomies and ontologies, are yet to be fully exploitedin a deep learning setting. This paper presents an efficient semantic text mining approach, whichconverts semantic information related to a given set of documents into a set of novel featuresthat are used for learning. The proposed Semantics-aware Recurrent deep Neural Architecture(SRNA) enables the system to learn simultaneously from the semantic vectors and from the raw textdocuments. We test the effectiveness of the approach on three text classification tasks: news topiccategorization, sentiment analysis and gender profiling. The experiments show that the proposedapproach outperforms the approach without semantic knowledge, with highest accuracy gain (up to10%) achieved on short document fragments.