Grounded Conversation Generation as Guided Traverses in Commonsense Knowledge Graphs
By Houyu Zhang, Zhenghao Liu, Chenyan Xiong, Zhiyuan Liu. 2019
158 citations recorded by the source corpus.
Human conversations naturally evolve around related concepts and hop to distant concepts. This paper presents a new conversation generation model, ConceptFlow, which leverages commonsense knowledge graphs to explicitly model conversation flows. By grounding conversations to the concept space, ConceptFlow represents the potential conversation flow as traverses in the concept space along commonsense relations. The traverse is guided by graph attentions in the concept graph, moving towards more meaningful directions in the concept space, in order to generate more semantic and informative responses. Experiments on Reddit conversations demonstrate ConceptFlow’s effectiveness over previous knowledge-aware conversation models and GPT-2 based models while using 70% fewer parameters, confirming the advantage of explicit modeling conversation structures. All source codes of this work are available at https://github.com/thunlp/ConceptFlow.
Subfields: Artificial Intelligence, Dialogue & Conversational AI, Knowledge Graphs & Representation
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Yejin Choi
- Xiang Ren
- Chenguang Zhu
- Mohit Bansal
- Percy Liang
- Jianfeng Gao
- Bill Yuchen Lin
- Hannaneh Hajishirzi
In Knowledge Graphs & Representation
- Xiang Ren
- Bill Yuchen Lin
- Yejin Choi
- Chandra Bhagavatula
- Chenyan Xiong
- Meng Jiang
- Wangchunshu Zhou
- Pei Zhou
Citations in the corpus
- CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning (2020, background): “, 2019b), and conversational systems (Zhang et al., 2020a).”
- Diversifying Content Generation for Commonsense Reasoning with Mixture of Knowledge Graph Experts (2022, background): “We note that as we targeted at the one-to-many generation problem, we excluded those baseline methods mentioned in the related work that cannot produce multiple outputs, e.g., Zhang et al. (2020a); Ji et al. (2020); Liu et al. (2021).”
- CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Society (2023, methodology): “There is also a large chunk of work that has proposed methods for automatic dataset creation [67, 57, 19, 75, 20, 98, 59, 96, 129, 62, 130, 86, 8].”
- Retrieval-Augmented Generation for AI-Generated Content: A Survey (2024, unknown): “Con-ceptFlow [234] leverages a commonsense knowledge graph to structure conversations, directing the flow of dialogue based on attention scores, and propelling the conversation forward.”
- ProtoQA: A Question Answering Dataset for Prototypical Common-Sense Reasoning (2020, unknown): “, 2019) and conversational text generation (Wu et al., 2020; Zhang et al., 2020).”
- Knowledge-Aware Language Model Pretraining (2020, background): “esearch mainly grounds the language model to some forms of explicit semantic structures, for example, additional entity-orientedtransformerlayers [16, 18], or explicitly stored knowledgegraph triples [19, 40, 41]. Groundingprovides a “safety net” for language models to look up a semantic facts in th”
- Recent advances in deep learning based dialogue systems: a systematic survey (2021, background): “Zhang et al. (2020) proposed LAtent BElief State (LABES) model, which treated the dialogue states as discrete latent variables to reduce the reliance on turn-level DST labels.”
- Knowledge-Augmented Methods for Natural Language Processing (2023, methodology): “…into language model with triplet information (Guan et al., 2020a); performing reasoning over knowledge graph via path finding strategies (Liu et al., 2019c; Ji et al., 2020a; Yu et al., 2022d); and improve the graph embeddings with graph neural networks (Zhang et al., 2020a; Ji et al., 2020b).”