S2SPMN: A Simple and Effective Framework for Response Generation with Relevant Information

Pei, Jiaxin; Li, Chenliang

doi:10.18653/v1/d18-1082

Cited by 12 publications

(10 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Xing et al (2017) asked whether the response can be used as a reply. Pei and Li (2018) asked whether the answer is natural for the question. Other studies have also evaluated the same or similar aspects by using keywords related to the connectivity, such as semantically appropriate for (Akama et al, 2017) or coherent with (Shen et al, 2017), and coherence (Lowe et al, 2017b).…”

Section: Criteria For Manual Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

Filtering Noisy Dialogue Corpora by Connectivity and Content Relatedness

Akama¹,

Yokoi

Suzuki

et al. 2020

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

View full text Add to dashboard Cite

Large-scale dialogue datasets have recently become available for training neural dialogue agents. However, these datasets have been reported to contain a non-negligible number of unacceptable utterance pairs. In this paper, we propose a method for scoring the quality of utterance pairs in terms of their connectivity and relatedness. The proposed scoring method is designed based on findings widely shared in the dialogue and linguistics research communities. We demonstrate that it has a relatively good correlation with the human judgment of dialogue quality. Furthermore, the method is applied to filter out potentially unacceptable utterance pairs from a large-scale noisy dialogue corpus to ensure its quality. We experimentally confirm that training data filtered by the proposed method improves the quality of neural dialogue agents in response generation. 1

show abstract

Section: Criteria For Manual Evaluationmentioning

confidence: 99%

“…Ritter et al (2011) suggested that an appropriate response should be on the same topic as the utterances. Several other studies have also focused on evaluating the relevance between an utterance and its response (Xu et al, 2018b;Pei and Li, 2018;Lowe et al, 2017b).…”

Section: Criteria For Manual Evaluationmentioning

confidence: 99%

Filtering Noisy Dialogue Corpora by Connectivity and Content Relatedness

Akama¹,

Yokoi

Suzuki

et al. 2020

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

View full text Add to dashboard Cite

show abstract

“…those which seek to generate responses or those which use a retrieval based approach. (Zhou et al, 2018;Pei and Li, 2018) are examples where the ability to generate responses is learnt from patterns in dialogues found in the training set. On the contrary, there exist several industrial scenarios where the domain is sufficiently restricted, or there exist legal ramifications associated with the responses, and hence pre-defined answers are preferred.…”

Section: Faq Assistantsmentioning

confidence: 99%

From Monolingual to Multilingual FAQ Assistant using Multilingual Co-training

Patidar¹,

Kumari²,

Patwardhan³

et al. 2019

Proceedings of the 2nd Workshop on Deep Learning Approaches for Low-Resource NLP (DeepLo 2019)

View full text Add to dashboard Cite

Recent research on cross-lingual transfershow state-of-the-art results on benchmark datasets using pre-trained language representation models (PLRM) like BERT. These results are achieved with the traditional training approaches, such as Zero-shot with no data, Translate-train or Translate-test with machine translated data. In this work, we propose an approach of "Multilingual Co-training" (MCT) where we augment the expert annotated dataset in the source language (English) with the corresponding machine translations in the target languages (e.g. Arabic, Spanish) and fine-tune the PLRM jointly. We observe that the proposed approach provides consistent gains in the performance of BERT for multiple benchmark datasets (e.g. 1.0% gain on MLDocs, and 1.2% gain on XNLI over translate-train with BERT), while requiring a single model for multiple languages. We further consider a FAQ dataset where the available English test dataset is translated by experts into Arabic and Spanish. On such a dataset, we observe an average gain of 4.9% over all other cross-lingual transfer protocols with BERT. We further observe that domainspecific joint pre-training of the PLRM using HR policy documents in English along with the machine translations in the target languages, followed by the joint finetuning, provides a further improvement of 2.8% in average accuracy.

show abstract

“…One significant task in conversation systems is response generation, which has attracted increasing attention from the academic community due to its ability to generate new responses that have not appeared in the corpus. Existing works on response generation include single-turn conversation [2]- [5] and multi-turn conversation [6], [7]. The former only uses the last utterance in context to generate a response, which contradicts the nature of real-world conversation.…”

Section: Introductionmentioning

confidence: 99%

A Hierarchical Structured Multi-Head Attention Network for Multi-Turn Response Generation

et al. 2020

View full text Add to dashboard Cite

As a crucial task in conversation systems, response generation for multi-turn conversation aims to generate a coherent, informative and diverse response according to the conversation context. Existing models of this task are limited in their ability to capture long-term dependencies within and between utterances and to identify pertinent or important information in the context. Inspired by the Transformer neural network based solely on attention mechanisms recently proposed in machine translation, we propose a novel hierarchical structured multi-head attention network (HMAN) model to address both problems. Specifically, the context sequences are encoded in a hierarchical structure in which multi-head self-attention is first employed to compute the representation matrix of each utterance, and then, these utterance matrices are integrated to form a context representation with the complicated dependencies learned using multi-head attention. The experimental results on two public conversation datasets demonstrate that our proposed model significantly outperforms several state-of-the-art baselines with respect to both automatic evaluation and human evaluation and can generate more diverse and informative responses.

show abstract

S2SPMN: A Simple and Effective Framework for Response Generation with Relevant Information

Cited by 12 publications

References 22 publications

Filtering Noisy Dialogue Corpora by Connectivity and Content Relatedness

Filtering Noisy Dialogue Corpora by Connectivity and Content Relatedness

From Monolingual to Multilingual FAQ Assistant using Multilingual Co-training

A Hierarchical Structured Multi-Head Attention Network for Multi-Turn Response Generation

Contact Info

Product

Resources

About