Exemplar Guided Neural Dialogue Generation

Cai, Hengyi; Chen, Hongshen; Song, Yujiang; Zhao, Xianfeng; Yin, Dawei

doi:10.24963/ijcai.2020/498

Cited by 8 publications

(8 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The methods similar to our instance-based generation are the skeleton-then-response frameworks which are popular in dialogue response generation (Cai et al, 2019;Wu et al, 2019;Yu and Jiang, 2021;Cai et al, 2020). These models usually treat the input text as a query and the similar query along with its response in databases is then retrieved with Information Retrieval (IR) systems.…”

Section: Retrieval-based Generationmentioning

confidence: 99%

DISK: Domain-constrained Instance Sketch for Math Word Problem Generation

Cao¹,

Zeng²,

Xu³

et al. 2022

Preprint

View full text Add to dashboard Cite

A math word problem (MWP) is a coherent narrative which reflects the underlying logic of math equations. Successful MWP generation can automate the writing of mathematics questions. Previous methods mainly generate MWP text based on inflexible pre-defined templates. In this paper, we propose a neural model for generating MWP text from math equations. Firstly, we incorporate a matching model conditioned on the domain knowledge to retrieve a MWP instance which is most consistent with the ground-truth, where the domain is a latent variable extracted with a domain summarizer. Secondly, by constructing a Quantity Cell Graph (QCG) from the retrieved MWP instance and reasoning over it, we improve the model's comprehension of real-world scenarios and derive a domainconstrained instance sketch to guide the generation. Besides, the QCG also interacts with the equation encoder to enhance the alignment between math tokens (e.g., quantities and variables) and MWP text. Experiments and empirical analysis on educational MWP set show that our model achieves impressive performance in both automatic evaluation metrics and human evaluation metrics. Equations: x = 6 * y ; (x + y) * 3= 147 Problem: Jane travels 6 times faster than mike. Traveling in opposite directions they are 147 miles apart after 3 hrs. Find their rates of travel. Equations: (1 -1/3 -9/20) * x = 245 Problem: At a local high school, 3/8 of the students are freshman. 1/4 are juniors. And 245 are seniors. Find the total number of students.

show abstract

Section: Retrieval-based Generationmentioning

confidence: 99%

DISK: Domain-constrained Instance Sketch for Math Word Problem Generation

Cao¹,

Zeng²,

Xu³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…To sidestep this issue, we adopt the idea of exemplars [5] for controllable text generation, where a set of sample responses from the training set, semantically related to the input context, is retrieved (using fine-tuned dense passage retrieval (DPR) [17] model) and fed to the response generator as templates [4]. These template responses guide the generator with stylistic and thematic cues on the response that are perceived as empathetic to the user.…”

Section: Mimementioning

confidence: 99%

“…This simple idea demonstrated good performance to the otherwise daunting task of response generation. Similar approaches were also explored in [4] where the retrieval process was designed to ensure both textual similarity and topical match with the generated response. While these approaches have been studied in generative applications like language modeling, open-domain dialog, or style transfer, to the best of our knowledge, we are the first to perform a systematic study of this paradigm for empathetic response generation.…”

Section: Introductionmentioning

confidence: 99%

Exemplars-guided Empathetic Response Generation Controlled by the Elements of Human Communication

Majumder¹,

Ghosal²,

Hazarika³

et al. 2021

Preprint

View full text Add to dashboard Cite

The majority of existing methods for empathetic response generation rely on the emotion of the context to generate empathetic responses. However, empathy is much more than generating responses with an appropriate emotion. It also often entails subtle expressions of understanding and personal resonance with the situation of the other interlocutor. Unfortunately, such qualities are difficult to quantify and the datasets lack the relevant annotations. To address this issue, in this paper we propose an approach that relies on exemplars to cue the generative model on fine stylistic properties that signal empathy to the interlocutor. To this end, we employ dense passage retrieval to extract relevant exemplary responses from the training set. Three elements of human communication-emotional presence, interpretation, and exploration-and sentiment are additionally introduced using synthetic labels to guide the generation towards empathy. The human evaluation is also extended by these elements of human communication. We empirically show that these approaches yield significant improvements in empathetic response quality in terms of both automated and human-evaluated metrics. The implementation is available at https://github.com/declare-lab/exemplary-empathy.

show abstract

“…Personalized content generation has attracted research interest in various domains, e.g., E-commerce (Zhao, Chen, and Yin 2019;Chen, Zhao, and Yin 2019), the automatic generation of marketing messages (Roy et al 2015;Chen et al 2020), persuasive message (Ding and Pan 2016;Zhang et al 2018), poetry generation (Shen, Guo, and Chen 2020), argument generation (Carenini and Moore 2006) and dialogue generation (Shen and Feng 2020;Feng et al 2020a;Shen, Feng, and Zhan 2019;Shen et al 2021;Cai et al 2020;Liu et al 2020). With the support of user preferences, the effectiveness has increases.…”

Section: Personalized Content Generationmentioning

confidence: 99%

Probing Product Description Generation via Posterior Distillation

Zhan¹,

Zhang²,

Chen³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

In product description generation (PDG), the user-cared aspect is critical for the recommendation system, which can not only improve user's experiences but also obtain more clicks. High-quality customer reviews can be considered as an ideal source to mine user-cared aspects. However, in reality, a large number of new products (known as long-tailed commodities) cannot gather sufficient amount of customer reviews, which brings a big challenge in the product description generation task. Existing works tend to generate the product description solely based on item information, i.e., product attributes or title words, which leads to tedious contents and cannot attract customers effectively. To tackle this problem, we propose an adaptive posterior network based on Transformer architecture that can utilize user-cared information from customer reviews. Specifically, we first extend the selfattentive Transformer encoder to encode product titles and attributes. Then, we apply an adaptive posterior distillation module to utilize useful review information, which integrates user-cared aspects to the generation process. Finally, we apply a Transformer-based decoding phase with copy mechanism to automatically generate the product description. Besides, we also collect a large-scare Chinese product description dataset to support our work and further research in this field. Experimental results show that our model is superior to traditional generative models in both automatic indicators and human evaluation.

show abstract

Exemplar Guided Neural Dialogue Generation

Cited by 8 publications

References 13 publications

DISK: Domain-constrained Instance Sketch for Math Word Problem Generation

DISK: Domain-constrained Instance Sketch for Math Word Problem Generation

Exemplars-guided Empathetic Response Generation Controlled by the Elements of Human Communication

Probing Product Description Generation via Posterior Distillation

Contact Info

Product

Resources

About