A Generative Attentional Neural Network Model for Dialogue Act Classification

Tran, Quan Hung; Haffari, Gholamreza; Zukerman, Ingrid

doi:10.18653/v1/p17-2083

Cited by 21 publications

(13 citation statements)

References 10 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…• ADC [47] and PDI [45] methds are acoustic and discourse classification based on HMM and SVM. • GAN [48] method is a generative neural network which incorporates attention technique and a label-to-label connection. Table III and Table IV…”

Section: Results and Analysismentioning

confidence: 99%

Improved Dynamic Memory Network for Dialogue Act Classification with Adversarial Training

Wan

Yan

Gao

et al. 2018

2018 IEEE International Conference on Big Data (Big Data)

View full text Add to dashboard Cite

Dialogue Act (DA) classification is a challenging problem in dialogue interpretation, which aims to attach semantic labels to utterances and characterize the speaker's intention. Currently, many existing approaches formulate the DA classification problem ranging from multi-classification to structured prediction, which suffer from two limitations: a) these methods are either handcrafted feature-based or have limited memories. b) adversarial examples can't be correctly classified by traditional training methods. To address these issues, in this paper we first cast the problem into a question and answering problem and proposed an improved dynamic memory networks with hierarchical pyramidal utterance encoder. Moreover, we apply adversarial training to train our proposed model. We evaluate our model on two public datasets, i.e., Switchboard dialogue act corpus and the MapTask corpus. Extensive experiments show that our proposed model is not only robust, but also achieves better performance when compared with some state-of-the-art baselines.

show abstract

Section: Results and Analysismentioning

confidence: 99%

Improved Dynamic Memory Network for Dialogue Act Classification with Adversarial Training

Wan

Yan

Gao

et al. 2018

2018 IEEE International Conference on Big Data (Big Data)

View full text Add to dashboard Cite

show abstract

“…Certainly, both paradigms are motivated by the sound reasoning that the constituent words, and their order within the sentence, are both key to interpreting its meaning, and hence both have been extensively explored within the literature. For example, Ahmadvand et al (2019), Liu and Lane (2017), Ortega and Vu (2017), Rojas-Barahona et al (2016), and Kalchbrenner and Blunsom (2013), all use variations of convolutional models as sentence encoders, while Li et al (2018), Papalampidi et al (2017), Tran et al (2017a) and Cerisara et al (2017), all employed recurrent architectures. Lee and Dernoncourt (2016) experimented with both convolutional and recurrent sentence encoders on several different corpora and found that neither approach was superior in all cases.…”

Section: Supervised Encodersmentioning

confidence: 99%

“…The authors do not define any training and test data split for the Maptask corpus; we randomly split the 128 dialogues into 3 parts. The training set comprises 80% of the dialogues (102), and the test and validation sets 10% each (13), which is similar to proportions used in previous studies (Tran et al 2017a;Tran et al 2017b).…”

Section: Maptaskmentioning

confidence: 99%

Sentence encoding for Dialogue Act classification

2021

View full text Add to dashboard Cite

In this study, we investigate the process of generating single-sentence representations for the purpose of Dialogue Act (DA) classification, including several aspects of text pre-processing and input representation which are often overlooked or underreported within the literature, for example, the number of words to keep in the vocabulary or input sequences. We assess each of these with respect to two DA-labelled corpora, using a range of supervised models, which represent those most frequently applied to the task. Additionally, we compare context-free word embedding models with that of transfer learning via pre-trained language models, including several based on the transformer architecture, such as Bidirectional Encoder Representations from Transformers (BERT) and XLNET, which have thus far not been widely explored for the DA classification task. Our findings indicate that these text pre-processing considerations do have a statistically significant effect on classification accuracy. Notably, we found that viable input sequence lengths, and vocabulary sizes, can be much smaller than is typically used in DA classification experiments, yielding no significant improvements beyond certain thresholds. We also show that in some cases the contextual sentence representations generated by language models do not reliably outperform supervised methods. Though BERT, and its derivative models, do represent a significant improvement over supervised approaches, and much of the previous work on DA classification.

show abstract

“…Lai et al (2019) introduce the Gated Self-Attention Memory Network (GSAMN). It combines gated attention (Dhingra et al, 2017;Tran et al, 2017), memory networks (Sukhbaatar et al, 2015) and self-attention (Vaswani et al, 2017) in one model. The authors use transfer learning with their Stack Exchange QA dataset.…”

Section: Related Workmentioning

confidence: 99%

Recursive Tree-Structured Self-Attention for Answer Sentence Selection

Mrini¹,

Farcas²,

Nakashole³

2021

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer

View full text Add to dashboard Cite

Syntactic structure is an important component of natural language text. Recent topperforming models in Answer Sentence Selection (AS2) use self-attention and transfer learning, but not syntactic structure. Tree structures have shown strong performance in tasks with sentence pair input like semantic relatedness. We investigate whether tree structures can boost performance in AS2. We introduce the Tree Aggregation Transformer: a novel recursive, tree-structured self-attention model for AS2. The recursive nature of our model is able to represent all levels of syntactic parse trees with only one additional self-attention layer. Without transfer learning, we establish a new state of the art on the popular TrecQA and WikiQA benchmark datasets. Additionally, we evaluate our method on four Community Question Answering datasets, and find that tree-structured representations have limitations with noisy user-generated text. We conduct probing experiments to evaluate how our models leverage tree structures across datasets. Our findings show that the ability of treestructured models to successfully absorb syntactic information is strongly correlated with a higher performance in AS2.

show abstract

A Generative Attentional Neural Network Model for Dialogue Act Classification

Cited by 21 publications

References 10 publications

Improved Dynamic Memory Network for Dialogue Act Classification with Adversarial Training

Improved Dynamic Memory Network for Dialogue Act Classification with Adversarial Training

Sentence encoding for Dialogue Act classification

Recursive Tree-Structured Self-Attention for Answer Sentence Selection

Contact Info

Product

Resources

About