Sequential Convolutional Neural Networks for Slot Filling in Spoken\n Language Understanding

2016/06/24 by Ngoc Thang Vu, Vu, Ngoc Thang
Computer Science · #Computation and Language (cs.CL) #FOS: Computer and information sciences #Natural Language Processing Techniques #Speech and dialogue systems #Topic Modeling

paper · pdf · doi:10.48550/arxiv.1606.07783

openalex publication_date 2016/06/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We investigate the usage of convolutional neural networks (CNNs) for the slot\nfilling task in spoken language understanding. We propose a novel CNN\narchitecture for sequence labeling which takes into account the previous\ncontext words with preserved order information and pays special attention to\nthe current word with its surrounding context. Moreover, it combines the\ninformation from the past and the future words for classification. Our proposed\nCNN architecture outperforms even the previously best ensembling recurrent\nneural network model and achieves state-of-the-art results with an F1-score of\n95.61% on the ATIS benchmark dataset without using any additional linguistic\nknowledge and resources.\n

Related

• Combining Word Feature Vector Method with the Convolutional Neural Network for Slot Filling in Spoken Language Understanding
• Effective Spoken Language Labeling with Deep Recurrent Neural Networks
• genCNN: A Convolutional Architecture for Word Sequence Prediction
• Exploiting Sentence and Context Representations in Deep Neural Models for Spoken Language Understanding
• Joint Learning of Word and Label Embeddings for Sequence Labelling in Spoken Language Understanding