Showing papers on "Surprise published in 2020"

Bias and Debias in Recommender System: A Survey and Future Directions

Abstract: While recent years have witnessed a rapid growth of research papers on recommender system (RS), most of the papers focus on inventing machine learning models to better fit user behavior data. However, user behavior data is observational rather than experimental. This makes various biases widely exist in the data, including but not limited to selection bias, position bias, exposure bias, and popularity bias. Blindly fitting the data without considering the inherent biases will result in many serious issues, e.g., the discrepancy between offline evaluation and online metrics, hurting user satisfaction and trust on the recommendation service, etc. To transform the large volume of research models into practical improvements, it is highly urgent to explore the impacts of the biases and perform debiasing when necessary. When reviewing the papers that consider biases in RS, we find that, to our surprise, the studies are rather fragmented and lack a systematic organization. The terminology "bias" is widely used in the literature, but its definition is usually vague and even inconsistent across papers. This motivates us to provide a systematic survey of existing work on RS biases. In this paper, we first summarize seven types of biases in recommendation, along with their definitions and characteristics. We then provide a taxonomy to position and organize the existing work on recommendation debiasing. Finally, we identify some open challenges and envision some future directions, with the hope of inspiring more research work on this important yet less investigated topic.

Surprise: A Python library for recommender systems

Abstract: Recommender systems aim at providing users with a list of recommendations of items that a service offers. For example, a video streaming service will typically rely on a recommender system to propose a personalized list of movies or series to each of its users. A typical problem in recommendation is that of rating prediction: given an incomplete dataset of useritem interations which take the form of numerical ratings (e.g. on a scale from 1 to 5), the goal is to predict the missing ratings for all remaining user-item pairs.

Fiscal Responses to COVID-19: Evidence from Local Governments and Nonprofits

Abstract: The rate of expansion and breadth of COVID-19 caught the world by surprise. From the perspective of nonprofit and public entities responsible for service provision, this pandemic is also unprecedented. We offer a RISE framework for navigating the fiscal effects of COVID-19 and rely on recent surveys to assess local governments' and nonprofit organizations' response strategies. We find that many nonprofits were hit the fastest and hardest by the pandemic and that local governments are, essentially, trying to figure out their financial condition moving into the next budget cycle.

Surprised-curious-confused: Epistemic emotions and knowledge exploration.

Abstract: Some epistemic emotions, such as surprise and curiosity, have attracted increasing scientific attention, whereas others, such as confusion, have yet to receive the attention they deserve. In addition, little is known about the relations between these emotions, their joint antecedents and outcomes, and how they differ from other emotions prompted during learning and knowledge generation (e.g., achievement emotions). In 3 studies (Ns = 102, 373, 125) using a trivia task with immediate feedback, we examined within-person interrelations, antecedents, and effects of 3 epistemic emotions (surprise, curiosity, and confusion). Studies 2 and 3 additionally included 2 achievement emotions (pride and shame). Using multilevel modeling to disentangle within- and between-person variance, we found that achievement emotions were associated with accuracy (i.e., correctness of the answer), whereas epistemic emotions were related to high-confidence errors (i.e., incorrect answers a person was confident in) generating cognitive incongruity. Furthermore, as compared with achievement emotions, epistemic emotions were more strongly and positively related to subsequent knowledge exploration. Specifically, surprise and curiosity were positive predictors of exploration. Confusion had positive predictive effects on exploration which were significant in Studies 1 and 3 but not in Study 2, suggesting that the effects of confusion are less stable and need to be investigated further. Apart from the findings for confusion, the results were fully robust across all 3 studies. They shed light on the distinct origins and outcomes of epistemic emotions. Directions for future research and practical implications are discussed. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

The Surprising Creativity of Digital Evolution: A Collection of Anecdotes from the Evolutionary Computation and Artificial Life Research Communities

Abstract: Evolution provides a creative fount of complex and subtle adaptations that often surprise the scientists who discover them. However, the creativity of evolution is not limited to the natural world: Artificial organisms evolving in computational environments have also elicited surprise and wonder from the researchers studying them. The process of evolution is an algorithmic process that transcends the substrate in which it occurs. Indeed, many researchers in the field of digital evolution can provide examples of how their evolving algorithms and organisms have creatively subverted their expectations or intentions, exposed unrecognized bugs in their code, produced unexpectedly adaptations, or engaged in behaviors and outcomes, uncannily convergent with ones found in nature. Such stories routinely reveal surprise and creativity by evolution in these digital worlds, but they rarely fit into the standard scientific narrative. Instead they are often treated as mere obstacles to be overcome, rather than results that warrant study in their own right. Bugs are fixed, experiments are refocused, and one-off surprises are collapsed into a single data point. The stories themselves are traded among researchers through oral tradition, but that mode of information transmission is inefficient and prone to error and outright loss. Moreover, the fact that these stories tend to be shared only among practitioners means that many natural scientists do not realize how interesting and lifelike digital organisms are and how natural their evolution can be. To our knowledge, no collection of such anecdotes has been published before. This article is the crowd-sourced product of researchers in the fields of artificial life and evolutionary computation who have provided first-hand accounts of such cases. It thus serves as a written, fact-checked collection of scientifically important and even entertaining stories. In doing so we also present here substantial evidence that the existence and importance of evolutionary surprises extends beyond the natural world, and may indeed be a universal property of all complex evolving systems.

Being surprised and surprising ourselves: A geography of personal and social change.

Abstract: Surprises are refuted expectations and therefore an inevitable concomitant of errors of anticipating the future. This paper argues that the timing is just right for a spatial account of surprise, or rather, for a geography of personal and social change that deploys the trope of surprise to help explain how and why change happens. Whether we are surprised by what transpires in our surroundings or we are surprising ourselves by leaping forward in impetuous deeds of reinventing who we are, the common denominator of these processes of becoming is that they produce geographical space and are produced by it.

Surprise, value and control in anterior cingulate cortex during speeded decision-making.

Abstract: Activity in the dorsal anterior cingulate cortex (dACC) is observed across a variety of contexts, and its function remains intensely debated in the field of cognitive neuroscience. While traditional views emphasize its role in inhibitory control (suppressing prepotent, incorrect actions), recent proposals suggest a more active role in motivated control (invigorating actions to obtain rewards). Lagging behind empirical findings, formal models of dACC function primarily focus on inhibitory control, highlighting surprise, choice difficulty and value of control as key computations. Although successful in explaining dACC involvement in inhibitory control, it remains unclear whether these mechanisms generalize to motivated control. In this study, we derive predictions from three prominent accounts of dACC and test these with functional magnetic resonance imaging during value-based decision-making under time pressure. We find that the single mechanism of surprise best accounts for activity in dACC during a task requiring response invigoration, suggesting surprise signalling as a shared driver of inhibitory and motivated control. The role of the anterior cingulate cortex (ACC) in decision-making and cognitive control is the subject of a long-standing debate. Vassena et al. tested the dominant accounts in the same paradigm and found that the ACC signals the difference between predicted and actual outcomes.

Behavioral, physiological, and neural signatures of surprise during naturalistic sports viewing

Abstract: Surprise signals a discrepancy between predicted and observed outcomes. It is theorized to segment the flow of experience into discrete perceived events, drive affective experiences, and create particularly resilient memories. However, the ability to precisely measure naturalistic surprise has remained elusive. We used advanced basketball analytics to derive a quantitative measure of surprise and characterized its behavioral, physiological, and neural effects on human subjects observing basketball games. We found that surprise served to segment ongoing experiences, as reflected in subjectively perceived event boundaries and shifts in neocortical neural patterns underlying belief states. Interestingly, these effects differed by whether surprising moments contradicted or bolstered current predominant beliefs. Surprise also positively correlated with pupil dilation, processing in subcortical regions associated with dopamine, game enjoyment, and, along with these physiological and neural measures, long-term memory. These investigations support key predictions from event segmentation theory and extend theoretical conceptualizations of surprise to real-world contexts.

What are We Depressed about When We Talk about COVID19: Mental Health Analysis on Tweets Using Natural Language Processing

Abstract: The outbreak of coronavirus disease 2019 (COVID-19) recently has affected human life to a great extent. Besides direct physical and economic threats, the pandemic also indirectly impact people's mental health conditions, which can be overwhelming but difficult to measure. The problem may come from various reasons such as unemployment status, stay-at-home policy, fear for the virus, and so forth. In this work, we focus on applying natural language processing (NLP) techniques to analyze tweets in terms of mental health. We trained deep models that classify each tweet into the following emotions: anger, anticipation, disgust, fear, joy, sadness, surprise and trust. We build the EmoCT (Emotion-Covid19-Tweet) dataset for the training purpose by manually labeling 1,000 English tweets. Furthermore, we propose and compare two methods to find out the reasons that are causing sadness and fear.

Emotional Analysis using Twitter Data during Pandemic Situation: COVID-19

Abstract: During the crisis situation caused due to COVID-19 disease, managing mental health and psychological well-being is as important as physical health of people. As web based life is broadly utilized by individuals to communicate their feeling and supposition, our framework utilizes Twitter information posted by individuals during this emergency circumstance to dissect the feelings of individuals. For processing the cleaned data NRC Word-Emotion Association Lexicon (have aka EmoLex) is used. NRC Word-Emotion Association Lexicon is a list of English with real-valued scores of intensity for eight basic emotion words ns (anger, anticipation, disgust, fear, joy, sadness, surprise, and trust). The text content of tweeter dataset created by fetching tweets across the world have classified into basic emotions like anger, anticipation, disgust, fear, joy, sadness, surprise and trust. This analysis can be used by authorities to understand the mental health of the people and can take necessary measures to decide on policies to fight against coronavirus which is affecting the social well-being as well as economy of the whole world.

An Integrated Real-Time Deep Learning and Belief Rule Base Intelligent System to Assess Facial Expression Under Uncertainty

Abstract: Nowadays, the recognition of facial expression draws significant attention in various domains. In view of this, a realtime facial expression recognition system has been developed using a Deep Learning approach, which can classify ten emotions, including angry, disgust, fear, happy, mockery, neutral, sad, surprise, think, and wink. In addition, an integrated expert system has also been developed by integrating Deep Learning with a Belief Rule Base to support the assessment of the overall mental state of a person over a period of time from video streaming data under uncertainty. In this research, data-driven and knowledge-driven approaches are integrated together to assess the mental state of an individual. Such a system could enable the identification of a suspect before committing any crime beforehand by the law enforcement agency. The performance of this integrated system is found reliable than existing methods of facial expression assessment. Contribution- The paper presents a noble method of computing the overall mental condition of a person by integrating CNN and BRBES under uncertainty. Contribution- The paper presents a noble method of computing the overall mental condition of a person by integrating CNN and BRBES under uncertainty.

Analyzing COVID-19 on Online Social Media: Trends, Sentiments and Emotions

Abstract: At the time of writing, the ongoing pandemic of coronavirus disease (COVID-19) has caused severe impacts on society, economy and people's daily lives. People constantly express their opinions on various aspects of the pandemic on social media, making user-generated content an important source for understanding public emotions and concerns. In this paper, we perform a comprehensive analysis on the affective trajectories of the American people and the Chinese people based on Twitter and Weibo posts between January 20th, 2020 and May 11th 2020. Specifically, by identifying people's sentiments, emotions (i.e., anger, disgust, fear, happiness, sadness, surprise) and the emotional triggers (e.g., what a user is angry/sad about) we are able to depict the dynamics of public affect in the time of COVID-19. By contrasting two very different countries, China and the Unites States, we reveal sharp differences in people's views on COVID-19 in different cultures. Our study provides a computational approach to unveiling public emotions and concerns on the pandemic in real-time, which would potentially help policy-makers better understand people's need and thus make optimal policy.

What are we depressed about when we talk about COVID-19: Mental health analysis on tweets using natural language processing

Abstract: The outbreak of coronavirus disease 2019 (COVID-19) recently has affected human life to a great extent. Besides direct physical and economic threats, the pandemic also indirectly impact people’s mental health conditions, which can be overwhelming but difficult to measure. The problem may come from various reasons such as unemployment status, stay-at-home policy, fear for the virus, and so forth. In this work, we focus on applying natural language processing (NLP) techniques to analyze tweets in terms of mental health. We trained deep models that classify each tweet into the following emotions: anger, anticipation, disgust, fear, joy, sadness, surprise and trust. We build the EmoCT (Emotion-Covid19-Tweet) dataset for the training purpose by manually labeling 1,000 English tweets. Furthermore, we propose an approach to find out the reasons that are causing sadness and fear, and study the emotion trend in both keyword and topic level.

Explanation-seeking curiosity in childhood

Abstract: Children are known for asking ‘why?’ — a query motivated by their desire for explanations. Research suggests that explanation-seeking curiosity (ESC) is triggered by first-person cues (such as novelty or surprise), third-person cues (such as a knowledgeable adults’ surprise or question), and future-oriented cues (such as expectations about information gain or future value). Once triggered, ESC is satisfied by an adequate explanation, typically obtained through causal intervention or question asking, both of which change in efficiency over development. ESC is an important driver of children’s learning because it combines the power of active learning and intrinsic motivation with the value of explanatory content, which can reveal the unobservable and causal structure of the world to support generalizable knowledge.

Examining the role of emotion in online consumer reviews of various attributes in the surprise Box shopping model

Abstract: The fierce competition among retailers demands the creation of new retail models. Emotional marketing aiming to arouse consumers' positive emotions attracts demand. In this context, the surprise box model, in which a business notifies consumers through a subscription and then mails boxes of new products without repetition, has emerged to rapidly develop. This study examines the role of consumer emotions in their online review writing behavior in the context of the surprise box shopping model. We find for the attributes of product, service, and fulfillment, but not for value; consumers tend to comment more in reviews when they have an extreme emotion, either positive or negative. Consumers comment even more when they have an extremely negative emotion than when they have an extremely positive emotion. In addition, we find that subscription period, experience, and consumers' overall satisfaction affect their review behavior, which depends on the particular attributes on which consumers comment. Further, we find that consumers' positive and negative emotions play different moderating roles in affecting the influence of the subscription period, subscription experience, and satisfaction on consumers' comments depending on the attribute types. This study helps firms understand how consumers perceive and evaluate various product and service attributes in the new surprise box business model. In this way, firms can better understand consumers' need to improve those attributes with different priorities. In addition, firms can use the positive electronic word-of-mouth effects generated from online consumer reviews to achieve both cognitive and affective empathy to attract future consumers.

Using persuasive refutation texts to prompt attitudinal and conceptual change

Abstract: We investigated knowledge and attitudes before and after reading refutation texts augmented by different kinds of persuasive information and how emotions mediated the process of knowledge and attitude change. Undergraduates (N = 424) enrolled in 4 universities from 3 countries read a refutation text on genetically modified foods (GMFs) and were then randomly assigned to receive additional information about advantages of GMFs, disadvantages of GMFs, or both. After studying, students reading about advantages of GMFs had significantly more positive attitudes than students who read about disadvantages. There was also a significant reduction in misconceptions; participants in the positive-oriented text condition showed the largest learning gains, particularly those who held more positive initial attitudes. Epistemic emotions of curiosity, frustration, hope, and enjoyment mediated attitude change while confusion mediated relations between prereading attitudes and postreading knowledge. In addition, the direct relationship between prior attitudes and surprise was moderated by type of text. When reading about both advantages and disadvantages of GMFs, participants reported significantly less surprise when compared with those who read about either advantages or disadvantages of GMFs. To foster conceptual change when learning about complex topics, refutation texts may be paired with persuasive information that is aligned with accurate conceptions. (PsycInfo Database Record (c) 2020 APA, all rights reserved)

COVID-19 – reflections on the surprise of both an expected and unexpected event

Abstract: Aim: This paper reflects on the COVID-19 epidemic from the perspective of small probabilities and the difficulty of predicting similar events. Against the background of basic economic principles, the importance of the precautionary principle for crisis management is discussed, as well as potential consequences of this epidemic. Findings: The authors argue that whilst the epidemic as such was unexpected, in future countries should be prepared for such stochastic events to happen. This requires a precautionary approach. When society is not prepared for such a calamity, or waits too long to implement measures to deal with it, the social and economic costs may be very high – much higher than ‘hedging bets’ and losing. The article reflects on different issues which are meant for further discussion on unpredictable future threats. One important issue is the uncertainty created by this event. This increases the likeliness that something unexpected can appear in the near future, creating the need for research and discussion on public and government responses to these events. Being aware of such challenges increases the likeliness of society and people to be prepared for such developments. It is concluded that the current crisis brings forward the question whether the current political-economic system and globalization makes future pandemics more likely, and whether a radical change towards a more locally oriented economy provides solutions that minimize the likelihood or frequency of future pandemics.

Dose-dependent modulation of the visually evoked N1/N170 by perceptual surprise: a clear demonstration of prediction-error signalling.

Abstract: Prediction-error checking processes play a key role in predictive coding models of perception. However, neural indices of such processes have yet to be unambiguously demonstrated. To date, experimental paradigms aiming to study such phenomena have relied upon the relative frequency of stimulus repeats and/or ‘unexpected’ events that are physically different from ‘expected’ events. These features of experimental design leave open alternative explanations for the observed effects. A definitive demonstration requires that presumed prediction error-related responses should show contextual dependency (rather than simply effects of frequency or repetition) and should not be attributable to low-level stimulus differences. Most importantly, prediction-error signals should show dose dependency with respect to the degree to which expectations are violated. Here, we exploit a novel experimental paradigm specifically designed to address these issues, using it to interrogate early latency event-related potentials (ERPs) to contextually expected and unexpected visual stimuli. In two electroencephalography (EEG) experiments, we demonstrate that an N1/N170 evoked potential is robustly modulated by unexpected perceptual events (‘perceptual surprise’) and shows dose-dependent sensitivity with respect to both the influence of prior information and the extent to which expectations are violated. This advances our understanding of perceptual predictions in the visual domain by clearly identifying these evoked potentials as an index of visual surprise.

Incorrect Provider Directories Associated With Out-Of-Network Mental Health Care And Outpatient Surprise Bills.

Abstract: Mental health services are up to six times more likely than general medical services to be delivered by an out-of-network provider, in part because many psychiatrists do not accept commercial insur...

Appraisal Theories for Emotion Classification in Text

Abstract: Automatic emotion categorization has been predominantly formulated as text classification in which textual units are assigned to an emotion from a predefined inventory, for instance following the fundamental emotion classes proposed by Paul Ekman (fear, joy, anger, disgust, sadness, surprise) or Robert Plutchik (adding trust, anticipation). This approach ignores existing psychological theories to some degree, which provide explanations regarding the perception of events. For instance, the description that somebody discovers a snake is associated with fear, based on the appraisal as being an unpleasant and non-controllable situation. This emotion reconstruction is even possible without having access to explicit reports of a subjective feeling (for instance expressing this with the words "I am afraid."). Automatic classification approaches therefore need to learn properties of events as latent variables (for instance that the uncertainty and the mental or physical effort associated with the encounter of a snake leads to fear). With this paper, we propose to make such interpretations of events explicit, following theories of cognitive appraisal of events, and show their potential for emotion classification when being encoded in classification models. Our results show that high quality appraisal dimension assignments in event descriptions lead to an improvement in the classification of discrete emotion categories. We make our corpus of appraisal-annotated emotion-associated event descriptions publicly available.

"Surprise" Out-of-network Billing in Orthopedic Surgery: Charges From Surprising Sources.

Abstract: I n 2019, lawmakers in both the Senate and House of Representatives proposed multiple bills to end surprise medical billing. A surprise medical bill is a charge from an out-of-network clinician that is not anticipated by the patient at the time of service. Several prior studies on surprise billing focus on emergency department and medical care, but less is known about the drivers of out-of-network surprise billing in elective surgical settings. Because surgical services require a multidisciplinary team of providers (eg, anesthesiologists, pathologists, etc), many of whom are not chosen by the patient and maybe out-of-network, patients receiving surgery may be particularly susceptible to surprise medical bills. We sought to determine the providers responsible for the highest rates and costs of out-of-network billing in orthopedic surgical episodes across the United States. We chose to study orthopedic procedures as they are among the most common and costly operations among commercially insured patients.

Unscripted drama: soccer audience response to suspense, surprise, and shock

Abstract: By modeling minute‐by‐minute television audience figures from English Premier League soccer matches, with close to 50,000 minute‐observations, we show that demand is partly driven by suspense and surprise. We also identify an additional relevant factor of appeal to audiences, namely shock, which refers to the difference between pre‐match and current game outcome probabilities. Suspense, surprise, and shock remain significant in the presence of a traditional measure of outcome uncertainty. (JEL C23, D12, L82, L83, Z20)

COVID-19 response and social work education in Nigeria: matters arising

Abstract: The novel COVID-19 pandemic has taken the world by surprise and given the scale at which it is impacting individuals, families, communities, and countries globally, the recognition of social worker...

Brain dynamics for confidence-weighted learning.

Abstract: Learning in a changing, uncertain environment is a difficult problem. A popular solution is to predict future observations and then use surprising outcomes to update those predictions. However, humans also have a sense of confidence that characterizes the precision of their predictions. Bayesian models use a confidence-weighting principle to regulate learning: for a given surprise, the update is smaller when the confidence about the prediction was higher. Prior behavioral evidence indicates that human learning adheres to this confidence-weighting principle. Here, we explored the human brain dynamics sub-tending the confidence-weighting of learning using magneto-encephalography (MEG). During our volatile probability learning task, subjects' confidence reports conformed with Bayesian inference. MEG revealed several stimulus-evoked brain responses whose amplitude reflected surprise, and some of them were further shaped by confidence: surprise amplified the stimulus-evoked response whereas confidence dampened it. Confidence about predictions also modulated several aspects of the brain state: pupil-linked arousal and beta-range (15-30 Hz) oscillations. The brain state in turn modulated specific stimulus-evoked surprise responses following the confidence-weighting principle. Our results thus indicate that there exist, in the human brain, signals reflecting surprise that are dampened by confidence in a way that is appropriate for learning according to Bayesian inference. They also suggest a mechanism for confidence-weighted learning: confidence about predictions would modulate intrinsic properties of the brain state to amplify or dampen surprise responses evoked by discrepant observations.

Decomposing neural responses to melodic surprise in musicians and non-musicians: evidence for a hierarchy of predictions in the auditory system

Abstract: Neural responses to auditory surprise are typically studied with highly unexpected, disruptive sounds. Consequently, little is known about auditory prediction in everyday contexts that are characterized by fine-grained, non-disruptive fluctuations of auditory surprise. To address this issue, we used IDyOM, a computational model of auditory expectation, to obtain continuous surprise estimates for a set of newly composed melodies. Our main goal was to assess whether the neural correlates of non-disruptive surprising sounds in a musical context are affected by musical expertise. Using magnetoencephalography (MEG), auditory responses were recorded from musicians and non-musicians while they listened to the melodies. Consistent with a previous study, the amplitude of the N1m component increased with higher levels of computationally estimated surprise. This effect, however, was not different between the two groups. Further analyses offered an explanation for this finding: Pitch interval size itself, rather than probabilistic prediction, was responsible for the modulation of the N1m, thus pointing to low-level sensory adaptation as the underlying mechanism. In turn, the formation of auditory regularities and proper probabilistic prediction were reflected in later components: the mismatch negativity (MMNm) and the P3am, respectively. Overall, our findings reveal a hierarchy of expectations in the auditory system and highlight the need to properly account for sensory adaptation in research addressing statistical learning. Highlights - In melodies, sound expectedness (modeled with IDyOM) is associated with the amplitude of the N1m. - This effect is not different between musicians and non-musicians. - Sensory adaptation related to melodic pitch intervals explains better the N1m effect. - Auditory regularities and the expectations captured by IDyOM are reflected in the MMNm and P3am. - Evidence for a hierarchy of auditory predictions during melodic listening.