Using Context Information for Dialog Act Classification in DNN Framework

Abstract

Previous work on dialog act (DA) classification has investigated different methods, such as hidden Markov models, maximum entropy, conditional random fields, graphical models, and support vector machines. A few recent studies explored using deep learning neural networks for DA classification, however, it is not clear yet what is the best method for using dialog context or DA sequential information, and how much gain it brings. This paper proposes several ways of using context information for DA classification, all in the deep learning framework. The baseline system classifies each utterance using the convolutional neural networks (CNN). Our proposed methods include using hierarchical models (recurrent neural networks (RNN) or CNN) for DA sequence tagging where the bottom layer takes the sentence CNN representation as input, concatenating predictions from the previous utterances with the CNN vector for classification, and performing sequence decoding based on the predictions from the sentence CNN model. We conduct thorough experiments and comparisons on the Switchboard corpus, demonstrate that incorporating context information significantly improves DA classification, and show that we achieve new state-of-the-art performance for this task.

Probabilistic Word Association for Dialogue Act Classification with Recurrent Neural Networks

“…To try and capture interactions between speakers, Kalchbrenner and Blunsom [10] used a Hierarchical CNN sentence model in conjunction with a RNN discourse model and condition the recurrent and output weights on the current speaker. Liu et al, [16] examine several different CNN and LSTM based architectures that incorporate different context information such as speaker change and dialogue history. Their work shows that including context information consistently yielded improvements over their baseline system.…”

“…This view has led much of the neural network based research to model the semantic content of a sentence, in conjunction with some other contextual information, such as previous utterance or DA sequences, or a change in speaker turn, to predict the appropriate DA for the current utterance [10,14]. Including such historical and contextual information has been shown to improve classification accuracy [16], and likely must be considered for any sophisticated classification model. However, motivated to examine the importance of different lexical and syntactic features, and their contribution towards DA classification, in this work each utterance is considered individually and out-of-context, that is, without any other contextual or historical information.…”

A post-processing method for detecting unknown intent of dialogue system via pre-trained deep neural network classifier

“…For modeling intents of multi-turn dialogue dataset, SwDA, we use hierarchical CNN (CNN+CNN) similar to the structure described in [23]. We illustrate the experiment flowchart in Figure 3.…”

“…However, there is no standard scheme for splitting SwDA dataset into training, validation and test sets. We follow the data splitting scheme suggested in [23] by randomly sampling 80% of conversation for the training set, 10% for the validation set and 10% for the test set. There are 162862 utterances for training, 20784 utterances for validation and 20146 utterances for testing.…”

Towards Universal Dialogue Act Tagging for Task-Oriented Dialogues

“…Our work represents short user utterances using recurrent neural networks, and additionally models dialogue context using a hierarchical recurrent neural network. Such dialogue-level models have also been proposed in [22] for dialogue act tagging of human-human social phone conversations. Previous studies mainly considered DA tagging of multi-human conversations, such as the Switchboard [15] corpus and meetings, such as the ICSI meeting corpus [23], whereas our focus lies on modeling system-side DAs.…”