Contextual Rephrase Detection for Reducing Friction in Dialogue Systems

Wang, Zhuoyi; Gupta, Saurabh; Hao, Jie; Fan, Xing; Li, Dingcheng; Li, Alexander Hanbo; Guo, Chenlei

doi:10.18653/v1/2021.emnlp-main.143

Cited by 5 publications

(3 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For re-ranking, given a pair of utterance and entity, we concatenate the output vector of CLS token of RoBERTa and the pooling output vector of GAT, and pass them to an MLP layer to produce the relevance score of the pair. For corrupt entity span detection, we predict the span's start and end positions at the token level, following similar approaches such as in [16] and [17]. Specifically, assume W S and W E are the start and the end vector respectively, and T i ∈ R H is the final hidden vector for the i th input token, then the score of a candidate span from position i to position j is computed as:…”

Section: L2 Re-ranking + Span Detectionmentioning

confidence: 99%

KG-ECO: Knowledge Graph Enhanced Entity Correction for Query Rewriting

Jinglun¹,

Li²,

Jiang³

et al. 2023

Preprint

View full text Add to dashboard Cite

Query Rewriting (QR) plays a critical role in large-scale dialogue systems for reducing frictions. When there is an entity error, it imposes extra challenges for a dialogue system to produce satisfactory responses. In this work, we propose KG-ECO: Knowledge Graph enhanced Entity COrrection for query rewriting, an entity correction system with corrupt entity span detection and entity retrieval/re-ranking functionalities.To boost the model performance, we incorporate Knowledge Graph (KG) to provide entity structural information (neighboring entities encoded by graph neural networks) and textual information (KG entity descriptions encoded by RoBERTa). Experimental results show that our approach yields a clear performance gain over two baselines: utterance level QR and entity correction without utilizing KG information. The proposed system is particularly effective for few-shot learning cases where target entities are rarely seen in training or there is a KG relation between the target entity and other contextual entities in the query.

show abstract

Section: L2 Re-ranking + Span Detectionmentioning

confidence: 99%

KG-ECO: Knowledge Graph Enhanced Entity Correction for Query Rewriting

Jinglun¹,

Li²,

Jiang³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…For textual data generation, Rik et al [23] used a Transformer to generate text from a knowledge graph. Zhuoyi et al [24] designed a novel rephrase detection system based on contextual content in dialogue scenes. Image generation and synthesis techniques are widely utilized.…”

Section: Attack and Defense In Other Layersmentioning

confidence: 99%

FakeSwarm: Improving Fake News Detection with Swarming Characteristics

Wu¹,

Ye²

2023

Natural Language Processing and Machine Learning

View full text Add to dashboard Cite

The proliferation of fake news poses a serious threat to society, as it can misinform and manipulate the public, erode trust in institutions, and undermine democratic processes. To address this issue, we present FakeSwarm, a fake news identification system that leverages the swarming characteristics of fake news. We propose a novel concept of fake news swarming characteristics and design three types of swarm features, including principal component analysis, metric representation, and position encoding, to extract the swarm behavior. We evaluate our system on a public dataset and demonstrate the effectiveness of incorporating swarm features in fake news identification, achieving an f1-score and accuracy over 97% by combining all three types of swarm features. Furthermore, we design an online learning pipeline based on the hypothesis of the temporal distribution pattern of fake news emergence, which is validated on a topic with early emerging fake news and a shortage of text samples, showing that swarm features can significantly improve recall rates in such cases. Our work provides a new perspective and approach to fake news detection and highlights the importance of considering swarming characteristics in detecting fake news.

show abstract

“…The information in these branches is fused to obtain a richer representation. Many other tasks [53] also use an attention mechanism to perform context modeling for better representations. For example, ContextNet [54] and Dual-mode ASR [55] that propose a novel CNN-RNN-transducer architecture with global context information for speech recognition, Cp-GAN for speech enhancement [56], and context aware attention in speech emotion detection [57].…”

mentioning

confidence: 99%

Data-driven Attention and Data-independent DCT based Global Context Modeling for Text-independent Speaker Recognition

Xia¹,

Hansen²

2022

Preprint

View full text Add to dashboard Cite

Learning an effective speaker representation is crucial for achieving reliable performance in speaker verification tasks. Speech signals are high-dimensional, long, and variablelength sequences that entail a complex hierarchical structure. Signals may contain diverse information at each time-frequency (TF) location. For example, it may be more beneficial to focus on highenergy parts for phoneme classes such as fricatives. The standard convolutional layer that operates on neighboring local regions cannot capture the complex TF global context information. In this study, a general global time-frequency context modeling framework is proposed to leverage the context information specifically for speaker representation modeling. First, a datadriven attention-based context model is introduced to capture the long-range and non-local relationship across different timefrequency locations. Second, a data-independent 2D-DCT based context model is proposed to improve model interpretability. A multi-DCT attention mechanism is presented to improve modeling power with alternate DCT base forms. Finally, the global context information is used to recalibrate salient timefrequency locations by computing the similarity between the global context and local features. The proposed lightweight blocks can be easily incorporated into a speaker model with little additional computational costs and effectively improves the speaker verification performance compared to the standard ResNet model and Squeeze&Excitation block by a large margin. Detailed ablation studies are also performed to analyze various factors that may impact performance of the proposed individual modules. Results from experiments show that the proposed global context modeling framework can efficiently improve the learned speaker representations by achieving channel-wise and timefrequency feature recalibration.

show abstract

Contextual Rephrase Detection for Reducing Friction in Dialogue Systems

Cited by 5 publications

References 10 publications

KG-ECO: Knowledge Graph Enhanced Entity Correction for Query Rewriting

KG-ECO: Knowledge Graph Enhanced Entity Correction for Query Rewriting

FakeSwarm: Improving Fake News Detection with Swarming Characteristics

Data-driven Attention and Data-independent DCT based Global Context Modeling for Text-independent Speaker Recognition

Contact Info

Product

Resources

About