Dual-View Conditional Variational Auto-Encoder for Emotional Dialogue Generation

2022 ◽  
Vol 21 (3) ◽  
pp. 1-18
Author(s):  
Mei Li ◽  
Jiajun Zhang ◽  
Xiang Lu ◽  
Chengqing Zong
Keyword(s):  
Relevant Information ◽  
Generation Process ◽  
Complementary Information ◽  
Emotional Information ◽  
Content Relevance ◽  
Latent Distribution ◽  
The Given

Emotional dialogue generation aims to generate appropriate responses that are content relevant with the query and emotion consistent with the given emotion tag. Previous work mainly focuses on incorporating emotion information into the sequence to sequence or conditional variational auto-encoder (CVAE) models, and they usually utilize the given emotion tag as a conditional feature to influence the response generation process. However, emotion tag as a feature cannot well guarantee the emotion consistency between the response and the given emotion tag. In this article, we propose a novel Dual-View CVAE model to explicitly model the content relevance and emotion consistency jointly. These two views gather the emotional information and the content-relevant information from the latent distribution of responses, respectively. We jointly model the dual-view via VAE to get richer and complementary information. Extensive experiments on both English and Chinese emotion dialogue datasets demonstrate the effectiveness of our proposed Dual-View CVAE model, which significantly outperforms the strong baseline models in both aspects of content relevance and emotion consistency.

scholarly journals Gas medium ionization and harmonic wavelength tunability in high-order harmonic generation with ultrashort laser pulses

2000 ◽  
Vol 18 (3) ◽  
pp. 477-482 ◽  
Author(s):  
P. CECCHERINI ◽  
A. BOSCOLO ◽  
L. POLETTO ◽  
G. TONDELLO ◽  
P. VILLORESI ◽  
...  
Keyword(s):  
Spectral Region ◽  
Blue Shift ◽  
Laser Pulses ◽  
High Order ◽  
Ultrashort Laser Pulses ◽  
Gas Jet ◽  
Generation Process ◽  
Free Electrons ◽  
High Order Harmonic ◽  
The Given

We have investigated the effect of free electrons on the spectral properties of high-order harmonics generated in a neon gas jet by a 30 fs Titanium:Sapphire pumping laser with intensities in the range 5–10 × 1014 W/cm2. The main feature of our observations concerns the possibility of continuously tuning the harmonic wavelength in the spectral region 20–7 nm, by taking advantage of the blue shift of harmonic wavelengths induced by the presence of free electrons to cover the entire spectral region between two consecutive harmonics of the unshifted spectrum. Different amounts of blue shift, which can be as large as 0.3–0.4 nm, are imparted to the given harmonic by simply changing the gas-jet-laser-beam-waist relative position. We have also interpreted our experimental results with a simple model for the generation process based on the “barrier suppression” ionization of an atom exposed to an ultraintense laser field.

Neural Oscillation Correlates Chemistry Decision-Making

2018 ◽  
Vol 28 (03) ◽  
pp. 1750031 ◽  
Author(s):  
Li-Yu Huang ◽  
Hsiao-Ching She ◽  
Tzyy-Ping Jung
Keyword(s):  
Decision Making ◽  
Undergraduate Students ◽  
Relevant Information ◽  
Stimulus Onset ◽  
Anterior Cingulate ◽  
Long Term Memory ◽  
Neural Oscillation ◽  
Beta Power ◽  
The Given

This study explored the electroencephalography (EEG) dynamics during a chemistry-related decision-making task and further examined whether the correctness of the decision-making performance could be reflected by EEG activity. A total of 66 undergraduate students’ EEG were collected while they participated in a chemistry-related decision-making task in which they had to retrieve the relevant chemistry concepts in order to make correct decisions for each task item. The results showed that it was only in the anterior cingulate cortex (ACC) cluster that distinct patterns in EEG dynamics were displayed for the correct and incorrect responses. The logistic regression results indicated that ACC theta power from 300[Formula: see text]ms to 250[Formula: see text]ms before stimulus onset was the most informative factor for estimating the likelihood of making correct decisions in the chemistry-related decision-making task, while it was the ACC low beta power from 150[Formula: see text]ms to 250[Formula: see text]ms after stimulus onset. The results suggested that the ACC theta augmentation before the stimulus onset serves to actively maintain the relevant information for retrieval from long-term memory, while the ACC low beta augmentation after the stimulus onset may serve the function of mapping the encoded stimulus onto the relevant criteria that the given participant has held within his or her mind to guide the decision-making responses.

scholarly journals ФИТОИНДИКАЦИОННОЕ ОЦЕНИВАНИЕ ИЗМЕРЕНИЙ, ПОЛУЧЕННЫХ ПРИ МНОГОМЕРНОМ ШКАЛИРОВАНИИ СТРУКТУРЫ РАСТИТЕЛЬНОГО СООБЩЕСТВА

2015 ◽  
Vol 5 (01) ◽  
pp. 69 ◽  
Author(s):  
A. V. Zhukov
Keyword(s):  
Multidimensional Scaling ◽  
Plant Community ◽  
Large Scale ◽  
Mantel Test ◽  
Specific Information ◽  
Scale Level ◽  
Complementary Information ◽  
Phenomenological Interpretation ◽  
Community Ordination ◽  
The Given

<p>The purpose of our work is to carry out plant community ordination by means of multidimensional scaling to reveal optimum ways of preliminary transformation of data and the similarity/dissimilarity measure, to identify multidimensional dimensions in terms of edafic properties and phytoindicator scales and to reveal character of interrelations of matrixes of plant community, phytoindicator scales and edafic properties. The received results testify that edafic and climatic scales matrixes bear the complementary information on edaphotop properties and possibly climatop. Most possibly that climatic scales at large-scale level bear the specific information on properties of environment. It is difficult to confirm, whether character of this information to adequate nominative properties of a scale at macrolevel is. But with confidence it is possible to say that climatic phytoindicator scales allow to differentiate ecological conditions in biogeocoenosis at large-scale level. Thus, at the given stage we tend to phenomenological interpretation of value of climatic phytoindicator scales at large-scale level.</p> <p><em>Keywords</em><em>: multidimensional scaling, community structure, phytoindicator scales, Mantel test</em></p>

Do children exploit the Maxim of Antecedent in order to interpret ambiguous descriptions?

1991 ◽  
Vol 18 (2) ◽  
pp. 451-457 ◽  
Author(s):  
Luca Surian
Keyword(s):  
Relevant Information ◽  
Definite Descriptions ◽  
The Other ◽  
Comprehension Strategy ◽  
Comprehension Task ◽  
The Given

ABSTRACTTwenty Italian six-year-olds and 20 eight-year-olds were asked to interpret eight ambiguous and eight clear definite descriptions. All ambiguous descriptions could refer to three drawings, one of which had been described by the subjects immediately before the comprehension task. In half of the trials with ambiguous messages the children's interlocutor was present while the children were describing the drawings; in the other half he was absent. In both conditions subjects showed a preference for the referents they had already described, indicating that they applied egocentrically a comprehension strategy based on the Maxim of Antecedent (Jackson & Jacobs, 1982). Children's failures to differentiate their responses in the two conditions are considered to be due to difficulties in taking account of the given-new distinction for relevant information.

scholarly journals Emotion Label Enhancement via Emotion Wheel and Lexicon

2021 ◽  
Vol 2021 ◽  
pp. 1-11
Author(s):  
Xueqiang Zeng ◽  
Qifan Chen ◽  
Sufen Chen ◽  
Jiali Zuo
Keyword(s):  
Distinct Advantage ◽  
Emotion Classification ◽  
Emotional Information ◽  
Basic Emotions ◽  
Enhancement Method ◽  
Emotion Label ◽  
Psychological Knowledge ◽  
Types Of Information ◽  
The Given ◽  
Emotion Labels

Emotion Distribution Learning (EDL) is a recently proposed multiemotion analysis paradigm, which identifies basic emotions with different degrees of expression in a sentence. Different from traditional methods, EDL quantitatively models the expression degree of the corresponding emotion on the given instance in an emotion distribution. However, emotion labels are crisp in most existing emotion datasets. To utilize traditional emotion datasets in EDL, label enhancement aims to convert logical emotion labels into emotion distributions. This paper proposed a novel label enhancement method, called Emotion Wheel and Lexicon-based emotion distribution Label Enhancement (EWLLE), utilizing the affective words’ linguistic emotional information and the psychological knowledge of Plutchik’s emotion wheel. The EWLLE method generates separate discrete Gaussian distributions for the emotion label of sentence and the emotion labels of sentiment words based on the psychological emotion distance and combines the two types of information into a unified emotion distribution by superposition of the distributions. The extensive experiments on 4 commonly used text emotion datasets showed that the proposed EWLLE method has a distinct advantage over the existing EDL label enhancement methods in the emotion classification task.

Geographical Map Annotation with Significant Tags available from Social Networks

2011 ◽  
pp. 425-448 ◽  
Author(s):  
Elena Roglia ◽  
Rosa Meo ◽  
Enrico Ponassi
Keyword(s):  
Social Networks ◽  
Web 2.0 ◽  
Statistical Tests ◽  
Geographical Area ◽  
Relevant Information ◽  
Personal Digital Assistants ◽  
Surrounding Areas ◽  
The Given ◽  
Selection Of

In this chapter we describe how to extract relevant information on a geographical area from information that users share and provide by means of their mobiles or personal digital assistants, thanks to Web 2.0 applications such as OpenStreetMap, Geonames, Flickr, and GoogleMaps. These Web 2.0 applications represent, store, and process information in an XML format. We analyze and use this information to enrich the content of the cartographic map of a given geographical area with up-to-date information. In addition we provide a characterization of the map by selection of the annotations that differentiate the given map from the surrounding areas. This occurs by means of statistical tests on the annotations frequency in the different geographical areas. We present the results of an experimental section in which we show that the content characterization is meaningful, statistically significant, and usefully concise.

Cognitive Evolution and the Transmission of Popular Narratives: A Literature Review and Application to Urban Legends

2018 ◽  
Vol 1 (1) ◽  
pp. 121 ◽  
Author(s):  
Joseph M. Stubbersfield ◽  
Emma G. Flynn ◽  
Jamshid J. Tehrani
Keyword(s):  
Content Analysis ◽  
Literature Review ◽  
Cultural Evolution ◽  
Cultural Transmission ◽  
Cognitive Biases ◽  
Relevant Information ◽  
Emotional Information ◽  
Selective Pressures ◽  
Urban Legends ◽  
Types Of Information

Recent research into cultural transmission suggests that humans are disposed to learn, remember, and transmit certain types of information more easily than others, and that any information that is passed between people will be subjected to cognitive selective pressures that alter the content and structure so as to make it maximally transmittable. This paper presents a review of emerging research on content biases in cultural evolution with relevance to the transmission of popular narratives. This is illustrated with content analysis of urban legends, which found that most exploited at least one known content bias, with emotional information and social information being the most frequent. We argue that the narratives do not succeed because of the transmission of adaptively relevant information but because of their exploitation of content biases in human cognition.Keywords: urban legends, content biases, cognitive biases, cultural evolution, cultural transmission

The Association between Technological Conditions and the Market Value of Equity

2008 ◽  
Vol 83 (2) ◽  
pp. 479-518 ◽  
Author(s):  
Zoltan P. Matolcsy ◽  
Anne Wyatt
Keyword(s):  
Technological Innovation ◽  
New Technologies ◽  
Technology Development ◽  
Investment Decisions ◽  
Market Value ◽  
Relevant Information ◽  
Generation Process ◽  
Earnings Growth ◽  
Future Earnings ◽  
Industry Research

The objective of this study is to provide evidence on how technological innovation conditions underlying the firm's investments drive earnings growth and, hence, market value of equity. Technologies develop and flourish or die out through the combined investment decisions of those firms doing the inventing, and those firms that adopt those inventions, and thereby help to spread (or diffuse) the innovations into wider use. Hence, technology is important for the investment decisions of all firms, regardless of whether they patent. We focus on three aggregate measures of technological innovation conditions: the success rate of past technological investments, technology complexity, and the technology development period. We use the interactions between each of these three conditions with earnings to capture the combined effect on market value of a firm's technological innovation environment. Our sample comprises 12,594 U.S. firm years for the period 1990–2000 including firms actively producing new technologies and firms that adopt technologies for their processes and products. Our primary and additional tests suggest that the interactions capture value-relevant information not reflected in commonly used variables including industry, research and development, sales, general, and administration expenses, risk, and growth. We also triangulate our results by providing evidence that aggregate technological innovation conditions predict future earnings and are, hence, instrumental in the earnings-generation process. This paper extends the valuation literature by (1) developing a generalizable framework that explains how technological innovation conditions link to future earnings and therefore map into the market value of equity; (2) developing aggregate measures of technological innovation conditions that are relevant for estimating future earnings and value for all firms; and (3) providing detailed empirical evidence on the relation between these aggregate measures and the market value of equity and earnings for all firms not just those that patent.

scholarly journals Mucko: Multi-Layer Cross-Modal Knowledge Reasoning for Fact-based Visual Question Answering

2020 ◽  
Author(s):  
Zihao Zhu ◽  
Jing Yu ◽  
Yujing Wang ◽  
Yajing Sun ◽  
Yue Hu ◽  
...  
Keyword(s):  
Question Answering ◽  
State Of The Art ◽  
Relevant Information ◽  
Convolutional Network ◽  
Graph Representations ◽  
Fine Grained ◽  
Visual Question Answering ◽  
Knowledge Reasoning ◽  
Final Answer ◽  
The Given

Fact-based Visual Question Answering (FVQA) requires external knowledge beyond the visible content to answer questions about an image. This ability is challenging but indispensable to achieve general VQA. One limitation of existing FVQA solutions is that they jointly embed all kinds of information without fine-grained selection, which introduces unexpected noises for reasoning the final answer. How to capture the question-oriented and information-complementary evidence remains a key challenge to solve the problem. In this paper, we depict an image by a multi-modal heterogeneous graph, which contains multiple layers of information corresponding to the visual, semantic and factual features. On top of the multi-layer graph representations, we propose a modality-aware heterogeneous graph convolutional network to capture evidence from different layers that is most relevant to the given question. Specifically, the intra-modal graph convolution selects evidence from each modality and cross-modal graph convolution aggregates relevant information across different graph layers. By stacking this process multiple times, our model performs iterative reasoning across three modalities and predicts the optimal answer by analyzing all question-oriented evidence. We achieve a new state-of-the-art performance on the FVQA task and demonstrate the effectiveness and interpretability of our model with extensive experiments.

Semantic Network Differences Across Engineering Design Communication Methods

2019 ◽  
Author(s):  
Christian E. Lopez ◽  
Zixuan V. Zhao ◽  
Conrad S. Tucker
Keyword(s):  
Semantic Network ◽  
Idea Generation ◽  
Relevant Information ◽  
Semantic Structure ◽  
Communication Effectiveness ◽  
Generation Process ◽  
Communication Methods ◽  
Product Idea ◽  
Virtual Studies ◽  
Selection Of

Abstract Engineering designers have a variety of methods at their disposal when it comes to communicating an idea (e.g., Linguistic, Pictorial, Virtual). Studies have explored how these methods affect the idea generation process, revealing that some methods can induce design fixation and reduce creativity. Moreover, studies reveal that depending on the communication methods and a receiver’s familiarity with the idea conveyed, the amount of relevant information transmitted could vary. Hence, based on previous studies, it is hypothesized that different communication methods and a receiver’s familiarity can impact a receiver’s ability to construct and interpret the information conveyed. To test this hypothesis, an experiment is conducted in which multiple methods are used to communicate different product ideas to individuals (N = 370). Participants are asked to describe the products in their own words and provide details about their functions. A text-mining approach is used to analyze the semantic structure of their responses. The results reveal that dissemination methods affect the consistency of participants’ responses, as well as the diversity of words used to describe a product idea or its functions. This knowledge can help designers in the selection of an appropriate method, given the design intention and help them leverage different methods to maximize communication effectiveness during the different stages of the design process.

Sign in / Sign up

Export Citation Format

Share Document