Augmenting End-to-End Dialogue Systems With Commonsense Knowledge

Young, Tom; Cambria, Erik; Chaturvedi, Iti; Zhou, Hao; Biswas, Subham; Huang, Minlie

doi:10.1609/aaai.v32i1.11923

Cited by 210 publications

(57 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Generally, the six baseline models can be divided into breadth-centric and depth-centric models. Tri-LSTM (Young et al 2018) is a breadth-centric model that augments its dialogue inputs with wide-ranging shallow KG facts to retrieve short KG paths. The other five baselines and HiTKG are depth-centric models which focus on a small set of KG entity-relation connections and perform deep reasoning over the KG.…”

Section: Experimental Settingsmentioning

confidence: 99%

“…Generating or retrieving responses according to the walking trajectory in KGs is effective in generating goal-oriented responses. The current graph walkers can generally be divided into recurrent walkers (Young et al 2018;Zhang et al 2019;Moon et al 2019;Li et al 2022) and graph attention based walkers (Jung, Son, and Lyu 2020). Recurrent walkers decode KG paths depending on a fixed-length vector, which creates a bottleneck for the performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

HiTKG: Towards Goal-Oriented Conversations via Multi-Hierarchy Learning

Ni¹,

Pandelea²,

Young³

et al. 2022

AAAI

Self Cite

View full text Add to dashboard Cite

Human conversations are guided by short-term and long-term goals. We study how to plan short-term goal sequences as coherently as humans do and naturally direct them to an assigned long-term goal in open-domain conversations. Goal sequences are a series of knowledge graph (KG) entity-relation connections generated by KG walkers that traverse through the KG. The existing recurrent and graph attention based KG walkers either insufficiently utilize the conversation states or lack global guidance. In our work, a hierarchical model learns goal planning in a hierarchical learning framework. We present HiTKG, a hierarchical transformer-based graph walker that leverages multiscale inputs to make precise and flexible predictions on KG paths. Furthermore, we propose a two-hierarchy learning framework that employs two stages to learn both turn-level (short-term) and global-level (long-term) conversation goals. Specifically, at the first stage, HiTKG is trained in a supervised fashion to learn how to plan turn-level goal sequences; at the second stage, HiTKG tries to naturally approach the assigned global goal via reinforcement learning. In addition, we propose MetaPath as the backbone method for KG path representation to exploit the entity and relation information concurrently. We further propose Multi-source Decoding Inputs and Output-level Length Head to improve the decoding controllability. Our experiments show that HiTKG achieves a significant improvement in the performance of turn-level goal learning compared with state-of-the-art baselines. Additionally, both automatic and human evaluation prove the effectiveness of the two-hierarchy learning framework for both short-term and long-term goal planning.

show abstract

Section: Experimental Settingsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

HiTKG: Towards Goal-Oriented Conversations via Multi-Hierarchy Learning

Ni¹,

Pandelea²,

Young³

et al. 2022

AAAI

Self Cite

View full text Add to dashboard Cite

show abstract

“…A related line of work (Liu et al, 2018;Moon et al, 2019;Xu et al, 2020;Young et al, 2018;Zhou et al, 2018a) that seemingly overcomes the aforementioned issues is knowledge selection from existing knowledge graphs (KGs) such as 2 Another example would be discourse relations between sentences, which we do not explore here.…”

Section: Introductionmentioning

confidence: 99%

Enhanced Knowledge Selection for Grounded Dialogues via Document Semantic Graphs

Li¹,

Namazifar²,

Jin³

et al. 2022

Preprint

View full text Add to dashboard Cite

Providing conversation models with background knowledge has been shown to make open-domain dialogues more informative and engaging. Existing models treat knowledge selection as a sentence ranking or classification problem where each sentence is handled individually, ignoring the internal semantic connection among sentences in background document. In this work, we propose to automatically convert the background knowledge documents into document semantic graphs and then perform knowledge selection over such graphs. Our document semantic graphs preserve sentence-level information through the use of sentence nodes and provide concept connections between sentences. We apply multitask learning for sentence-level knowledge selection and concept-level knowledge selection jointly, and show that it improves sentencelevel selection. Our experiments show that our semantic graph based knowledge selection improves over sentence selection baselines for both the knowledge selection task and the end-to-end response generation task on HollE (Moghe et al., 2018) and improves generalization on unseen topics in WoW (Dinan et al., 2019). 1

show abstract

“…In reality, all such knowledge cannot be prepared in advance. A knowledge base is not only necessary for providing various services such as information search and recommendation, but also effective for non-task-oriented dialogue systems in order to prevent generic or dull responses (Xing et al, 2017;Young et al, 2018;Zhou et al, 2018;. However, it is impractical to presuppose a perfect knowledge base (West et al, 2014).…”

Section: Introductionmentioning

confidence: 99%

Lexical Acquisition during Dialogues through Implicit Confirmation

Komatani

Ono

Takeda

et al. 2022

dad

View full text Add to dashboard Cite

We have been addressing the problem of acquiring attributes of unknown terms through dialogues and previously proposed an approach using the implicit confirmation process. It is crucial for dialogue systems to ask questions that do not diminish the user’s willingness to talk. In this paper, we conducted a user study to investigate user impression for several question types, including explicit and implicit, to acquire lexical knowledge. We clarified the order among the types and found that repeating the same question type annoys the user and degrades user impression even when the content of the questions is correct. We also propose a method for determining whether an estimated attribute is correct, which is included in an implicit question. The method exploits multiple-user responses to implicit questions about the attribute of the same unknown term. Experimental results revealed that the proposed method exhibited a higher precision rate for determining the correctly estimated attributes than when only single-user responses were considered.

show abstract

Augmenting End-to-End Dialogue Systems With Commonsense Knowledge

Cited by 210 publications

References 22 publications

HiTKG: Towards Goal-Oriented Conversations via Multi-Hierarchy Learning

HiTKG: Towards Goal-Oriented Conversations via Multi-Hierarchy Learning

Enhanced Knowledge Selection for Grounded Dialogues via Document Semantic Graphs

Lexical Acquisition during Dialogues through Implicit Confirmation

Contact Info

Product

Resources

About