Showing papers by "Peter A. Gloor published in 2021"
TL;DR: Patients perceive positively teams where doctors take the leadership of the communication network and ensure an effective team conversation, and patient satisfaction and service perceptions are greatly influenced by behavioral and network factors.
20 citations
TL;DR: The results reveal that the proposed approaches for classifying emotions from speech by combining conventional mel-frequency cepstral coefficients (MFCCs) with image features extracted from spectrograms by a pretrained convolutional neural network (CNN) lead to an improvement in prediction accuracy.
Abstract: Speech is one of the most natural communication channels for expressing human emotions. Therefore, speech emotion recognition (SER) has been an active area of research with an extensive range of applications that can be found in several domains, such as biomedical diagnostics in healthcare and human–machine interactions. Recent works in SER have been focused on end-to-end deep neural networks (DNNs). However, the scarcity of emotion-labeled speech datasets inhibits the full potential of training a deep network from scratch. In this paper, we propose new approaches for classifying emotions from speech by combining conventional mel-frequency cepstral coefficients (MFCCs) with image features extracted from spectrograms by a pretrained convolutional neural network (CNN). Unlike prior studies that employ end-to-end DNNs, our methods eliminate the resource-intensive network training process. By using the best prediction model obtained, we also build an SER application that predicts emotions in real time. Among the proposed methods, the hybrid feature set fed into a support vector machine (SVM) achieves an accuracy of 0.713 in a 6-class prediction problem evaluated on the Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) dataset, which is higher than the previously published results. Interestingly, MFCCs taken as unique input into a long short-term memory (LSTM) network achieve a slightly higher accuracy of 0.735. Our results reveal that the proposed approaches lead to an improvement in prediction accuracy. The empirical findings also demonstrate the effectiveness of using a pretrained CNN as an automatic feature extractor for the task of emotion prediction. Moreover, the success of the MFCC-LSTM model is evidence that, despite being conventional features, MFCCs can still outperform more sophisticated deep-learning feature sets.
14 citations
TL;DR: In this article, the authors combine social network and semantic analysis to identify top performers based on email communication and find that top performers tend to assume central network positions and have high responsiveness to emails, and use more positive and complex language with low emotionality, but rich in influential words that are probably reused by co-workers.
Abstract: In the information economy, individuals' work performance is closely associated with their digital communication strategies. This study combines social network and semantic analysis to develop a method to identify top performers based on email communication. By reviewing existing literature, we identified the indicators that quantify email communication into measurable dimensions. To empirically examine the predictive power of the proposed indicators, we collected 2 million email archive of 578 executives in an international service company. Panel regression was employed to derive interpretable association between email indicators and top performance. The results suggest that top performers tend to assume central network positions and have high responsiveness to emails. In email contents, top performers use more positive and complex language, with low emotionality, but rich in influential words that are probably reused by co-workers. To better explore the predictive power of the email indicators, we employed AdaBoost machine learning models, which achieved 83.56% accuracy in identifying top performers. With cluster analysis, we further find three categories of top performers, "networkers" with central network positions, "influencers" with influential ideas and "positivists" with positive sentiments. The findings suggest that top performers have distinctive email communication patterns, laying the foundation for grounding email communication competence in theory. The proposed email analysis method also provides a tool to evaluate the different types of individual communication styles.
13 citations
01 May 2021
TL;DR: The Happimeter as mentioned in this paper was used over three months in the innovation lab of a bank with 22 employees to measure individual happiness, activity, and stress, and participants were randomly divided into an experimental and a control group of similar size.
Abstract: Happiness has been an overarching goal of mankind at least since Aristotle spoke of Eudaimonia. However measuring happiness has been elusive and until now has almost exclusively been done by asking survey questions about self-perceived happiness. We propose a novel approach, tracking happiness and stress through changes in body signals with a smartwatch, the “Happimeter”. It predicts individual emotions from sensor data collected by an Android Wear smartwatch, such as acceleration, heartbeat, and activity. The Happimeter was used over three months in the innovation lab of a bank with 22 employees to measure individual happiness, activity, and stress. The participants were randomly divided into an experimental and a control group of similar size. Both groups wore the watch and entered their subjective happiness, activity and stress levels several times a day. The user-entered ratings were then used to train a machine learning system using the sensors of the smartwatch to subsequently automatically predict happiness, activity, and stress. The experimental group received ongoing feedback about their mood and which activity, sensor signals, or other people, made them happier or unhappier, while the control group did not get any feedback about their predicted and manually entered emotions. Just like in quantum physics we observed a “Heisenberg Effect”, where the participants made aware of their measurements changed their behavior: Members of the experimental group that received happiness feedback were 16% happier, and 26% more active than the control group at the end of the experiment. No effect was observed for stress.
12 citations
TL;DR: In this article, the influence of emotions of meeting participants on the perceived outcome of video meetings was studied, and it was found that the presentations that triggered wide swings in “fear” and “joy” among the participants are correlated with a higher rating.
Abstract: In the last 14 months, COVID-19 made face-to-face meetings impossible and this has led to rapid growth in videoconferencing. As highly social creatures, humans strive for direct interpersonal interaction, which means that in most of these video meetings the webcam is switched on and people are “looking each other in the eyes”. However, it is far from clear what the psychological consequences of this shift to virtual face-to-face communication are and if there are methods to alleviate “videoconferencing fatigue”. We have studied the influence of emotions of meeting participants on the perceived outcome of video meetings. Our experimental setting consisted of 35 participants collaborating in eight teams over Zoom in a one semester course on Collaborative Innovation Networks in bi-weekly video meetings, where each team presented its progress. Emotion was tracked through Zoom face video snapshots using facial emotion recognition that recognized six emotions (happy, sad, fear, anger, neutral, and surprise). Our dependent variable was a score given after each presentation by all participants except the presenter. We found that the happier the speaker is, the happier and less neutral the audience is. More importantly, we found that the presentations that triggered wide swings in “fear” and “joy” among the participants are correlated with a higher rating. Our findings provide valuable input for online video presenters on how to conduct better and less tiring meetings; this will lead to a decrease in “videoconferencing fatigue”.
8 citations
TL;DR: In this paper, the authors used social network analysis and text mining to measure publication attributes and understand which variables can better help predicting their future success, and they were able to predict scholarly impact of a paper in terms of citations received six years after publication with almost 80 percent accuracy.
Abstract: Citations acknowledge the impact a scientific publication has on subsequent work. At the same time, deciding how and when to cite a paper, is also heavily influenced by social factors. In this work, we conduct an empirical analysis based on a dataset of 2010-2012 global publications in chemical engineering. We use social network analysis and text mining to measure publication attributes and understand which variables can better help predicting their future success. Controlling for intrinsic quality of a publication and for the number of authors in the byline, we are able to predict scholarly impact of a paper in terms of citations received six years after publication with almost 80 percent accuracy. Results suggest that, all other things being equal, it is better to co-publish with rotating co-authors and write the papers' abstract using more positive words, and a more complex, thus more informative, language. Publications that result from the collaboration of different social groups also attract more citations.
7 citations
TL;DR: In this paper, the authors used a region-based convolutional neural network (RNN) to detect equine emotions based on established behavioral ethograms indicating emotional affect through the head, neck, ear, muzzle, and eye position.
Abstract: Creating intelligent systems capable of recognizing emotions is a difficult task, especially when looking at emotions in animals. This paper describes the process of designing a “proof of concept” system to recognize emotions in horses. This system is formed by two elements, a detector and a model. The detector is a fast region-based convolutional neural network that detects horses in an image. The model is a convolutional neural network that predicts the emotions of those horses. These two elements were trained with multiple images of horses until they achieved high accuracy in their tasks. In total, 400 images of horses were collected and labeled to train both the detector and the model while 40 were used to test the system. Once the two components were validated, they were combined into a testable system that would detect equine emotions based on established behavioral ethograms indicating emotional affect through the head, neck, ear, muzzle, and eye position. The system showed an accuracy of 80% on the validation set and 65% on the test set, demonstrating that it is possible to predict emotions in animals using autonomous intelligent systems. Such a system has multiple applications including further studies in the growing field of animal emotions as well as in the veterinary field to determine the physical welfare of horses or other livestock.
6 citations
Posted Content•
TL;DR: A literature review focusing on the use of Natural Language Generation (NLG) to automatically detect and generate persuasive texts is presented in this paper, focusing on generative aspects through conceptualizing determinants of persuasion in five business-focused categories: benevolence, linguistic appropriacy, logical argumentation, trustworthiness, tools and datasets.
Abstract: This literature review focuses on the use of Natural Language Generation (NLG) to automatically detect and generate persuasive texts. Extending previous research on automatic identification of persuasion in text, we concentrate on generative aspects through conceptualizing determinants of persuasion in five business-focused categories: benevolence, linguistic appropriacy, logical argumentation, trustworthiness, tools and datasets. These allow NLG to increase an existing message's persuasiveness. Previous research illustrates key aspects in each of the above mentioned five categories. A research agenda to further study persuasive NLG is developed. The review includes analysis of seventy-seven articles, outlining the existing body of knowledge and showing the steady progress in this research field.
2 citations
TL;DR: Through extensive simulations on both synthetic and real-world networks, it is found that for networks with theoretically infinite size, only a finite and small number of sensor nodes are needed to detect the global spreading information with almost certainty.
Abstract: As a powerful and low-cost instant information dissemination platform, large-scale online social networks (OSNs) play a pivotal role in shaping our modern information age. The efficient detection of wide-spreading information in OSNs is very important in many aspects including public opinion supervision, social governance, stock markets, counter-terrorism and presidential election. However, real-world OSNs have gigantic sizes and thus their full structural data are usually unavailable, making this problem extremely challenging. In this work, we illustrate the close mapping between efficient detection and optimal spreading from the perspective of network percolation theory. This analogy inspires us to propose a theory of using only limited local network information to select the optimal set of information sensors. Through extensive simulations on both synthetic and real-world networks, we find that for networks with theoretically infinite size, only a finite and small number of sensor nodes are needed to detect the global spreading information with almost certainty. Most importantly, we empirically confirm the utility of our theory on the largest micro blog in China by crawling almost the full Sina Weibo social network with 99,546,027 users in 2014 and the real spreading data of Weibo messages.
1 citations
TL;DR: In this paper, the authors compared risk-taking attitudes assessed with the Domain Specific Risk-Taking (DOSPERT) survey with individual e-mail networking patterns and body language measured with smartwatches.
Abstract: As the Enron scandal and Bernie Madoff’s pyramid scheme have shown, individuals’ attitude towards ethical risks can have a huge impact on society at large. In this paper, we compare risk-taking attitudes assessed with the Domain-Specific Risk-Taking (DOSPERT) survey with individual e-mail networking patterns and body language measured with smartwatches. We find that e-mail communication signals such as network structure and dynamics, and content features as well as real-world behavioral signals measured through a smartwatch such as heart rate, acceleration, and mood state demonstrate a strong correlation with the individuals’ risk-preference in the different domains of the DOSPERT survey. For instance, we found that people with higher degree centrality in the e-mail network show higher likelihood to take social risks, while using language expressing a “you live only once” attitude indicates lower willingness to take risks in some domains. Our results show that analyzing the human interaction in organizational networks provides valuable information for decision makers and managers to support an increase in ethical behavior of the organization’s members.
1 citations
Posted Content•
TL;DR: The authors conducted case studies to identify adaptations that professional and non-professional writers make in written scenarios to increase their subjective persuasiveness, and identified challenges that those writers faced and identified strategies to resolve them with persuasive natural language generation, i.e., artificial intelligence.
Abstract: The ability to persuade others is critical to professional and personal success. However, crafting persuasive messages is demanding and poses various challenges. We conducted nine exploratory case studies to identify adaptations that professional and non-professional writers make in written scenarios to increase their subjective persuasiveness. Furthermore, we identified challenges that those writers faced and identified strategies to resolve them with persuasive natural language generation, i.e., artificial intelligence. Our findings show that humans can achieve high degrees of persuasiveness (more so for professional-level writers), and artificial intelligence can complement them to achieve increased celerity and alignment in the process.
Posted Content•
TL;DR: In this paper, the authors used a region-based convolutional neural network (RNN) to detect equine emotions based on established behavioral ethograms indicating emotional affect through head, neck, ear, muzzle, and eye position.
Abstract: Creating intelligent systems capable of recognizing emotions is a difficult task, especially when looking at emotions in animals. This paper describes the process of designing a "proof of concept" system to recognize emotions in horses. This system is formed by two elements, a detector and a model. The detector is a faster region-based convolutional neural network that detects horses in an image. The second one, the model, is a convolutional neural network that predicts the emotion of those horses. These two models were trained with multiple images of horses until they achieved high accuracy in their tasks, creating therefore the desired system. 400 images of horses were used to train both the detector and the model while 80 were used to validate the system. Once the two components were validated they were combined into a testable system that would detect equine emotions based on established behavioral ethograms indicating emotional affect through head, neck, ear, muzzle, and eye position. The system showed an accuracy of between 69% and 74% on the validation set, demonstrating that it is possible to predict emotions in animals using autonomous intelligent systems. It is a first "proof of concept" approach that can be enhanced in many ways. Such a system has multiple applications including further studies in the growing field of animal emotions as well as in the veterinary field to determine the physical welfare of horses or other livestock.
TL;DR: This work presents a deep learning forecasting framework which is capable to predict tomorrow's news topics on Twitter and news feeds based on yesterday's content and topic-interaction features and demonstrates that selecting a topic’s historical local neighbors in the topic-network as extra features greatly improves the prediction accuracy and outperforms existing baselines.
Abstract: Rapid advances in machine learning combined with wide availability of online social media have created considerable research activity in predicting what might be the news of tomorrow based on an analysis of the past. In this work, we present a deep learning forecasting framework which is capable to predict tomorrow's news topics on Twitter and news feeds based on yesterday's content and topic-interaction features. The proposed framework starts by generating topics from words using word embeddings and K-means clustering. Then temporal topic-networks are constructed where two topics are linked if the same user has worked on both topics. Structural and dynamic metrics calculated from networks along with content features and past activity, are used as input of a long short-term memory (LSTM) model, which predicts the number of mentions of a specific topic on the subsequent day. Utilizing dependencies among topics, our experiments on two Twitter datasets and the HuffPost news dataset demonstrate that selecting a topic's historical local neighbors in the topic-network as extra features greatly improves the prediction accuracy and outperforms existing baselines.