Automatic generation of sentimental texts via mixture adversarial networks

Wang, Ke; Wan, Xiaojun

doi:10.1016/j.artint.2019.07.003

Cited by 27 publications

(18 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, gene-wise expression states were first inferred by the left truncated mixture Gaussian model. Modules of genes that show consistent activated or suppressed expression in a subset of cells were identified using a non-negative matrix factorization method namely MEBF [ 65 ]. We further evaluated the enrichment of the genes in each module against known targets of transcriptional regulatory factors, and the association of the cells of each module with the experimental conditions.…”

Section: Resultsmentioning

confidence: 99%

Ref-1 redox activity alters cancer cell metabolism in pancreatic cancer: exploiting this novel finding as a potential target

Gampala

Shah

et al. 2021

J Exp Clin Cancer Res

View full text Add to dashboard Cite

Background Pancreatic cancer is a complex disease with a desmoplastic stroma, extreme hypoxia, and inherent resistance to therapy. Understanding the signaling and adaptive response of such an aggressive cancer is key to making advances in therapeutic efficacy. Redox factor-1 (Ref-1), a redox signaling protein, regulates the conversion of several transcription factors (TFs), including HIF-1α, STAT3 and NFκB from an oxidized to reduced state leading to enhancement of their DNA binding. In our previously published work, knockdown of Ref-1 under normoxia resulted in altered gene expression patterns on pathways including EIF2, protein kinase A, and mTOR. In this study, single cell RNA sequencing (scRNA-seq) and proteomics were used to explore the effects of Ref-1 on metabolic pathways under hypoxia. Methods scRNA-seq comparing pancreatic cancer cells expressing less than 20% of the Ref-1 protein was analyzed using left truncated mixture Gaussian model and validated using proteomics and qRT-PCR. The identified Ref-1’s role in mitochondrial function was confirmed using mitochondrial function assays, qRT-PCR, western blotting and NADP assay. Further, the effect of Ref-1 redox function inhibition against pancreatic cancer metabolism was assayed using 3D co-culture in vitro and xenograft studies in vivo. Results Distinct transcriptional variation in central metabolism, cell cycle, apoptosis, immune response, and genes downstream of a series of signaling pathways and transcriptional regulatory factors were identified in Ref-1 knockdown vs Scrambled control from the scRNA-seq data. Mitochondrial DEG subsets downregulated with Ref-1 knockdown were significantly reduced following Ref-1 redox inhibition and more dramatically in combination with Devimistat in vitro. Mitochondrial function assays demonstrated that Ref-1 knockdown and Ref-1 redox signaling inhibition decreased utilization of TCA cycle substrates and slowed the growth of pancreatic cancer co-culture spheroids. In Ref-1 knockdown cells, a higher flux rate of NADP + consuming reactions was observed suggesting the less availability of NADP + and a higher level of oxidative stress in these cells. In vivo xenograft studies demonstrated that tumor reduction was potent with Ref-1 redox inhibitor similar to Devimistat. Conclusion Ref-1 redox signaling inhibition conclusively alters cancer cell metabolism by causing TCA cycle dysfunction while also reducing the pancreatic tumor growth in vitro as well as in vivo.

show abstract

Section: Resultsmentioning

confidence: 99%

Ref-1 redox activity alters cancer cell metabolism in pancreatic cancer: exploiting this novel finding as a potential target

Gampala

Shah

et al. 2021

J Exp Clin Cancer Res

View full text Add to dashboard Cite

show abstract

“…Models for Emotional Dialogs Previous work on emotional dialogs fall into three categories: (1) controlled emotional dialog generation , Colombo et al, 2019, Song et al, 2019, (Zhou and Wang, 2018 1. The two large-scale automatic annotated dataset NLPCC2017 and MOJITALK (Zhou and Wang, 2018) and the manually labeled dataset DailyDialog (Li et al, 2017) are widely used for controlled emotional dialog generation , Zhou and Wang, 2018, Wang and Wan, 2019, Shen and Feng, 2020. The Empatheticdialog (Rashkin et al, 2019) dataset is designed for training empathetic dialog models (Lin et al, 2019, Majumder et al, 2020, Li et al, 2020a.…”

Section: Related Workmentioning

confidence: 99%

Towards Multi-Turn Empathetic Dialogs with Positive Emotion Elicitation

Wang¹,

Xu²,

Wu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Emotional support is a crucial skill for many real-world scenarios, including caring for the elderly, mental health support, and customer service chats. This paper presents a novel task of empathetic dialog generation with positive emotion elicitation to promote users' positive emotion, similar to that of emotional support between humans. In this task, the agent conducts empathetic responses along with the target of eliciting the user's positive emotions in the multi-turn dialog. To facilitate the study of this task, we collect a large-scale emotional dialog dataset with positive emotion elicitation, called PosEmoDial (about 820k dialogs, 3M utterances). In these dialogs, the agent tries to guide the user from any possible initial emotional state, e.g., sadness, to a positive emotional state. Then we present a positiveemotion-guided dialog generation model with a novel loss function design. This loss function encourages the dialog model to not only elicit positive emotions from users but also ensure smooth emotional transitions along with the whole dialog. Finally, we establish benchmark results on PosEmoDial, and we will release this dataset and related source code to facilitate future studies. * The work was done when Shihang Wang was doing internship at Baidu.

show abstract

“…When added to the original set of training data, such kinds of counterfactually augmented data (CAD) have shown their benefits on learning real causal associations and improving the model robustness in recent studies (Kaushik et al, 2020(Kaushik et al, , 2021Wang and Culotta, 2021). Unlike gradient-based adversarial examples (Wang and Wan, 2019;Zang et al, 2020), which cannot provide a clear boundary between positive and negative instances to humans, counterfactuals could provide "human-like" logic to show a modification to the arXiv:2106.15231v2 [cs.CL] 30 Jun 2021 input that makes a difference to the output classification (Byrne, 2019).…”

Section: Introductionmentioning

confidence: 99%

Exploring the Efficacy of Automatically Generated Counterfactuals for Sentiment Analysis

Yang¹,

Li²,

Cunningham³

et al. 2021

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer

View full text Add to dashboard Cite

While state-of-the-art NLP models have been achieving the excellent performance of a wide range of tasks in recent years, important questions are being raised about their robustness and their underlying sensitivity to systematic biases that may exist in their training and test data. Such issues come to be manifest in performance problems when faced with out-ofdistribution data in the field. One recent solution has been to use counterfactually augmented datasets in order to reduce any reliance on spurious patterns that may exist in the original data. Producing high-quality augmented data can be costly and time-consuming as it usually needs to involve human feedback and crowdsourcing efforts. In this work, we propose an alternative by describing and evaluating an approach to automatically generating counterfactual data for the purpose of data augmentation and explanation. A comprehensive evaluation on several different datasets and using a variety of state-of-the-art benchmarks demonstrate how our approach can achieve significant improvements in model performance when compared to models training on the original data and even when compared to models trained with the benefit of human-generated augmented data.

show abstract

Automatic generation of sentimental texts via mixture adversarial networks

Cited by 27 publications

References 8 publications

Ref-1 redox activity alters cancer cell metabolism in pancreatic cancer: exploiting this novel finding as a potential target

Ref-1 redox activity alters cancer cell metabolism in pancreatic cancer: exploiting this novel finding as a potential target

Towards Multi-Turn Empathetic Dialogs with Positive Emotion Elicitation

Exploring the Efficacy of Automatically Generated Counterfactuals for Sentiment Analysis

Contact Info

Product

Resources

About