What You Say and How You Say it: Joint Modeling of Topics and Discourse in Microblog Conversations

  title={What You Say and How You Say it: Joint Modeling of Topics and Discourse in Microblog Conversations},
  author={Jichuan Zeng and Jing Li and Yulan He and Cuiyun Gao and Michael R. Lyu and Irwin King},
  journal={Transactions of the Association for Computational Linguistics},
This paper presents an unsupervised framework for jointly modeling topic content and discourse behavior in microblog conversations. Concretely, we propose a neural model to discover word clusters indicating what a conversation concerns (i.e., topics) and those reflecting how participants voice their opinions (i.e., discourse).1 Extensive experiments show that our model can yield both coherent topics and meaningful discourse behavior. Further study shows that our topic and discourse… 

Who Responded to Whom: The Joint Effects of Latent Topics and Discourse in Conversation Structure

A model is proposed to learn latent topics and discourse in word distributions, and predict pairwise initiationresponse links via exploiting topic consistency and discourse dependency, and significantly outperforms the previous state of the arts.

Continuity of Topic, Interaction, and Query: Learning to Quote in Online Conversations

This work studies automatic quotation generation in an online conversation and explores how language consistency affects whether a quotation fits the given context and captures the contextual consistency of a quotation in terms of latent topics, interactions with the dialogue history, and coherence to the query’s existing contents.

Modeling Global and Local Interactions for Online Conversation Recommendation

A novel framework to automatically recommend conversations to users based on what they said and how they behaved in their chatting histories is presented, which significantly outperforms the state-of-the-art model.

Is Discourse Role Important for Emotion Recognition in Conversation?

A novel method to exploit latent discourse role information of an utterance to determine the emotion it conveys in a conversation is proposed, using a variant of the Variational-Autoencoder to model the context-aware latent discourse roles of each utterance in an unsupervised way.

DialogBERT: Discourse-Aware Response Generation via Learning to Recover and Rank Utterances

Experiments show that this approach remarkably outperforms three baselines, such as BART and DialoGPT, in terms of quantitative evaluation, and the human evaluation suggests that DialogBERT generates more coherent, informative, and human-like responses than the baselines with significant margins.

What Changed Your Mind: The Roles of Dynamic Topics and Discourse in Argumentation Process

A novel neural model is proposed that is able to dynamically track the changes of latent topics and discourse in argumentative conversations, allowing the investigation of their roles in influencing the outcomes of persuasion.

Topic-Aware Neural Keyphrase Generation for Social Media Language

This work proposes a sequence-to-sequence (seq2seq) based neural keyphrase generation framework, enabling absent keyphrases to be created, and shows that the model significantly outperforms both extraction and generation models without exploiting latent topics.

Multi-view learning for context-aware extractive summarization

A novel context-aware extractive summarization algorithm is proposed by utilizing the concept of multi-view learning, which achieves the state-of-the-art summarization results when compared with other contemporary approaches.

Improving Context Modeling in Neural Topic Segmentation

A segmenter based on a hierarchical attention BiLSTM network is enhanced by adding a coherence-related auxiliary task and restricted self-attention to better model context, and outperforms SOTA approaches when trained and tested on three datasets.



A Joint Model of Conversational Discourse Latent Topics on Microblogs

An unsupervised model that jointly learns word distributions to represent different roles of conversational discourse and various latent topics in reflecting content information can effectively indicate summary-worthy content in microblog conversations.

Microblog Conversation Recommendation via Joint Modeling of Topics and Discourse

A statistical model is proposed that jointly captures topics for representing user interests and conversation content, and discourse modes for describing user replying behavior and conversation dynamics that outperforms methods that only model content without considering discourse.

Joint Modeling of Content and Discourse Relations in Dialogues

Experimental results show that the joint model can outperform state-of-the-art approaches for both phrase-based content selection and discourse relation prediction tasks and on predicting the consistency among team members' understanding of their group decisions.

Topic Extraction from Microblog Posts Using Conversation Structures

This work organize microblog posts as conversation trees based on reposting and replying relations, which enrich context information to alleviate data sparseness and captures the different extents that leader and follower messages may contain the key topical words, thus further enhances the quality of the induced topics.

You Talking to Me? A Corpus and Algorithm for Conversation Disentanglement

This work proposes a graph-theoretic model for disentanglement, using discourse-based features which have not been previously applied to this task, and presents a corpus of Internet Relay Chat dialogue in which the various conversations have been manually disentangled.

Find the Conversation Killers: A Predictive Study of Thread-ending Posts

This study is particularly interested in identifying a post in a multi-party conversation that is unlikely to be further replied to, which therefore kills that thread of the conversation, and proposes a deep learning model called the ConverNet.

Unsupervised Modeling of Twitter Conversations

This work proposes the first unsupervised approach to the problem of modeling dialogue acts in an open domain, trained on a corpus of noisy Twitter conversations, and addresses the challenge of evaluating the emergent model with a qualitative visualization and an intrinsic conversation ordering task.

Improving LDA topic models for microblogs via tweet pooling and automatic labeling

This paper empirically establishes that a novel method of tweet pooling by hashtags leads to a vast improvement in a variety of measures for topic coherence across three diverse Twitter datasets in comparison to an unmodified LDA baseline and a range of pooling schemes.

A biterm topic model for short texts

The approach can discover more prominent and coherent topics, and significantly outperform baseline methods on several evaluation metrics, and is found that BTM can outperform LDA even on normal texts, showing the potential generality and wider usage of the new topic model.

Unsupervised Modeling of Dialog Acts in Asynchronous Conversations

A graph-theoretic deterministic framework and two probabilistic conversation models are investigated for modeling dialog acts in emails and forums and empirical evaluation suggests HMM+Mix is a better conversation model than the simple HMM model.