Sound Natural: Content Rephrasing in Dialog Systems

Einolghozati, Arash; Gupta, Anchit; Diedrick, Keith; Gupta, Sonal

doi:10.18653/v1/2020.emnlp-main.414

Cited by 4 publications

(6 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, to ensure the grammatical correctness of all the generated questions and answers, we include two rounds of a peer-review process. Similar to [8], the first round included a set of annotators where we asked them to assert the produced results and rephrase them if needed. This step ensures more natural and fluent dialogues.…”

Section: Results Validationmentioning

confidence: 99%

“…We substitute the tokens back to the original position after the first version is generated. Similarly to other works [8], we use box brackets to distinguish the seed answer from the remaining sentence; this is helpful when experimenting with the verbalized answers. For example, for the question "What countries did the main character travel in the book Eat, Pray, Love?"…”

Section: Initial Answer Verbalizationmentioning

confidence: 99%

“…However, existing open-source KGQA systems are restricted to only generating or producing answers without verbalizing them in natural language [10]. The lack of verbalization makes the interaction with the user unnatural and often leaves the users with ambiguity [8]. Let us consider the first question in Figure 1, "What countries did the main character travel in the book Eat, Pray, Love?…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

An Answer Verbalization Dataset for Conversational Question Answerings over Knowledge Graphs

Kacupaj¹,

Singh²,

Maleshkova³

et al. 2022

Preprint

View full text Add to dashboard Cite

We introduce a new dataset for conversational question answering over Knowledge Graphs (KGs) with verbalized answers. Question answering over KGs is currently focused on answer generation for single-turn questions (KGQA) or multiple-tun conversational question answering (ConvQA). However, in a real-world scenario (e.g., voice assistants such as Siri, Alexa, and Google Assistant), users prefer verbalized answers. This paper contributes to the state-of-the-art by extending an existing ConvQA dataset with multiple paraphrased verbalized answers. We perform experiments with five sequence-to-sequence models on generating answer responses while maintaining grammatical correctness. We additionally perform an error analysis that details the rates of models' mispredictions in specified categories. Our proposed dataset (Verbal-ConvQuestions) extended with answer verbalization is publicly available with detailed documentation on its usage for wider utility 1 . CCS CONCEPTS• Information systems → Question answering.

show abstract

Section: Results Validationmentioning

confidence: 99%

Section: Initial Answer Verbalizationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

An Answer Verbalization Dataset for Conversational Question Answerings over Knowledge Graphs

Kacupaj¹,

Singh²,

Maleshkova³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…We add a copy pointer head to allow AnswerBART to copy tokens directly from the query and the passages. We follow Einolghozati et al (2020) which first initializes the copy head with the average of the last layer's pretrained decoder attention head, and adds a loss that forces the decoder to use the copying mechanism.…”

Section: Model Extensionsmentioning

confidence: 99%

Conversational Answer Generation and Factuality for Reading Comprehension Question-Answering

Peshterliev,

Oguz,

Chatterjee

et al. 2021

Preprint

View full text Add to dashboard Cite

Question answering (QA) is an important use case on voice assistants. A popular approach to QA is extractive reading comprehension (RC) which finds an answer span in a text passage. However, extractive answers are often unnatural in a conversational context which results in suboptimal user experience. In this work, we investigate conversational answer generation for QA. We propose Answer-BART, an end-to-end generative RC model which combines answer generation from multiple passages with passage ranking and answerability. Moreover, a hurdle in applying generative RC are hallucinations where the answer is factually inconsistent with the passage text. We leverage recent work from summarization to evaluate factuality. Experiments show that AnswerBART significantly improves over previous best published results on MS MARCO 2.1 NLGEN by 2.5 ROUGE-L and Narra-tiveQA by 9.4 ROUGE-L.

show abstract

“…While their goal is very similar to ours, rephrasing transcripts is very different from rephrasing emails, as we show in Section 5. Einolghozati et al (2020) applied a pre-trained BART with a copy mechanism for the task of rephrasing virtual assistance messages. However, they are focused on style adaptation and personal pronouns modification while we focus on contextbased enrichment.…”

Section: Related Workmentioning

confidence: 99%

Automatic Rephrasing of Transcripts-based Action Items

Cohen¹,

Kantor²,

Hilleli³

et al. 2021

Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021

View full text Add to dashboard Cite

The automated transcription of spoken language, and meetings, in particular, is becoming more widespread as automatic speech recognition systems are becoming more accurate. This trend has significantly accelerated since the outbreak of the COVID-19 pandemic, which led to a major increase in the number of online meetings. However, the transcription of spoken language has not received much attention from the NLP community compared to documents and other forms of written language. In this paper, we study a variation of the summarization problem over the transcription of spoken language: given a transcribed meeting, and an action item (i.e., a commitment or request to perform a task), our goal is to generate a coherent and self-contained rephrasing of the action item. To this end, we compiled a novel dataset of annotated meeting transcripts, including human rephrasing of action items. We use state-of-the-art supervised text generation techniques and establish a strong baseline based on BART and UniLM (two pretrained transformer models). Due to the nature of natural speech, language is often broken and incomplete and the task is shown to be harder than an analogous task over email data. Particularly, we show that the baseline models can be greatly improved once models are provided with additional information. We compare two approaches: one incorporating features extracted by coreferenceresolution. Additional annotations are used to train an auxiliary model to detect the relevant context in the text. Based on the systematic human evaluation, our best models exhibit nearhuman-level rephrasing capability on a constrained subset of the problem.

show abstract

Sound Natural: Content Rephrasing in Dialog Systems

Cited by 4 publications

References 17 publications

An Answer Verbalization Dataset for Conversational Question Answerings over Knowledge Graphs

An Answer Verbalization Dataset for Conversational Question Answerings over Knowledge Graphs

Conversational Answer Generation and Factuality for Reading Comprehension Question-Answering

Automatic Rephrasing of Transcripts-based Action Items

Contact Info

Product

Resources

About