Crowd Phenomena Research Articles

Modeling emergency department nursing workload in real time: An exploratory study

Study of emergency department (ED) nursing workload has been largely subsumed under the related but separate phenomenon of ED crowding. Nursing workload is difficult to quantify directly. This observational study explored modeling ED nursing workload indirectly, in real time, from quantitative data available from the patient tracking computer system (PTCS). MethodsData on 2793 patient visits plus departmental statistics were collected during 167 60-minute survey periods (SP) in a 25-bed hospital ED in the United States. The charge nurse assessed a perceived workload score (WLS) according to pre-determined criteria following each SP as a validation measure. Data analysisCorrelations were calculated between the data and WLS, and strongly correlating variables were incorporated into linear regression models that sought to approximate WLS. ResultsA measure of aggregate patient acuity derived from the Emergency Severity Index (ESI) was the strongest predictor of WLS (r = 0.7991). The best-performing model agreed with WLS in 64% of SPs. ConclusionsGood agreement between model output and WLS suggests that ED nursing workload can be estimated indirectly in real time using data from a PTCS. Strong correlation between the ESI derivative and WLS further validates ESI and suggests a new application for the ESI score.

What Do Employees Know? Evidence from a Social Media Platform

ABSTRACT We use employee predictions of their companies' six-month business outlook from Glassdoor.com to assess the information content of employee social media disclosures. We find that average employee outlook is incrementally informative in predicting future operating performance. Its information content is greater when the disclosures are aggregated from a larger, more diverse, more knowledgeable employee base, consistent with the wisdom of crowds phenomenon. Average outlook predicts bad news events more strongly than good news events, suggesting that employee social media disclosures are relatively more important as a source of bad news. Consistent with the organizational theory, we find systematic differences in the quantity and nature of the information in employee disclosures when the disclosures are grouped based on employee attributes and job responsibilities. Finally, average outlook predicts future returns of firms that attract less attention by analysts and investors, suggesting that investors in these firms use outlook inefficiently.

What Do Employees Know? Evidence from a Social Media Platform

We use employee predictions of their companies’ six-month business outlook from Glassdoor.com to assess the information content of employee social media disclosures. We find that average employee outlook is incrementally informative in predicting future operating performance. Its information content is greater when the disclosures are aggregated from a larger, more diverse, more knowledgeable employee base, consistent with the wisdom of crowds phenomenon. Average outlook predicts bad news events more strongly than good news events, suggesting that employee social media disclosures are relatively more important as a source of bad news. Consistent with the organizational theory, we find systematic differences in the quantity and nature of the information in employee disclosures when the disclosures are grouped based on employee attributes and job responsibilities. Finally, average outlook predicts future returns of firms that attract less attention by analysts and investors, suggesting that investors in these firms use outlook inefficiently.

Wisdom of Microcrowds in Evaluating Solutions to Esoteric Engineering Problems

A multitude of studies in economics, psychology, political and social sciences have demonstrated the wisdom of crowds (WoC) phenomenon, where the collective estimate of a group can be more accurate than estimates of individuals. While WoC is observable in such domains where the participating individuals have experience or familiarity with the question at hand, it remains unclear how effective WoC is for domains that traditionally require deep expertise or sophisticated computational models to estimate objective answers. This work explores how effective WoC is for engineering design problems that are esoteric in nature, that is, problems (1) whose solutions traditionally require expertise and specialized knowledge, (2) where access to experts can be costly or infeasible, and (3) in which previous WoC studies with the general population have been shown to be highly ineffective. The main hypothesis in this work is that in the absence of experts, WoC can be observed in groups that consist of practitioners who are defined to have a base familiarity with the problems in question but not necessarily domain experts. As a way to emulate commonly encountered engineering problem-solving scenarios, this work studies WoC with practitioners that form microcrowds consisting of 5–15 individuals, thereby giving rise to the term the wisdom of microcrowds (WoMC). Our studies on design evaluations show that WoMC produces results whose mean is in the 80th percentile or better across varying crowd sizes, even for problems that are highly nonintuitive in nature.

Convolutional Nonlinear Differential Recurrent Neural Networks for Crowd Scene Understanding

With the growth of crowd phenomena in the real world, crowd scene understanding is becoming an important task in anomaly detection and public security. Visual ambiguities and occlusions, high density, low mobility, and scene semantics, however, make this problem a great challenge. In this paper, we propose an end-to-end deep architecture, convolutional nonlinear differential recurrent neural networks (CNDRNNs), for crowd scene understanding. CNDRNNs consist of GoogleNet Inception V3 convolutional neural networks (CNNs) and nonlinear differential recurrent neural networks (RNNs). Different from traditional non-end-to-end solutions which separate the steps of feature extraction and parameter learning, CNDRNN utilizes a unified deep model to optimize the parameters of CNN and RNN hand in hand. It thus has the potential of generating a more harmonious model. The proposed architecture takes sequential raw image data as input, and does not rely on tracklet or trajectory detection. It thus has clear advantages over the traditional flow-based and trajectory-based methods, especially in challenging crowd scenarios of high density and low mobility. Taking advantage of CNN and RNN, CNDRNN can effectively analyze the crowd semantics. Specifically, CNN is good at modeling the semantic crowd scene information. On the other hand, nonlinear differential RNN models the motion information. The individual and increasing orders of derivative of states (DoS) in differential RNN can progressively build up the ability of the long short-term memory (LSTM) gates to detect different levels of salient dynamical patterns in deeper stacked layers modeling higher orders of DoS. Lastly, existing LSTM-based crowd scene solutions explore deep temporal information and are claimed to be “deep in time.” Our proposed method CNDRNN, however, models the spatial and temporal information in a unified architecture and achieves “deep in space and time.” Extensive performance studies on the Violent-Flows, CUHK Crowd, and NUS-HGA datasets show that the proposed technique significantly outperforms state-of-the-art methods.

Trolls, bans and reverts: simulating Wikipedia

The surprisingly high reliability of Wikipedia has often been seen as a beneficial effect of the aggregation of diverse contributors, or as an instance of the wisdom of crowds phenomenon; additional factors such as elite contributors, Wikipedia’s policy or its administration have also been mentioned. We adjudicate between such explanations by modelling and simulating the evolution of a Wikipedia entry. The main threat to Wikipedia’s reliability, namely the presence of epistemically disruptive agents such as disinformers and trolls, turns out to be offset only by a combination of factors: Wikipedia’s administration and the possibility to instantly revert entries, both of which are insufficient when considered in isolation. Our results suggest that the reliability of Wikipedia should receive a pluralist explanation, involving factors of different kinds.

Multiview-based group behavior analysis in optical image sequence

Group behavior analysis is a hot topic in intelligent video surveillance, and has attracted a surge of interest in the field of artificial intelligence. Groups are the basic components of a crowd system, and provide a high-level representation of the crowd phenomenon. By investigating the motion dynamics within each image patch, this paper proposes a multiview-based group behavior analysis method that is able to divide the paths into different groups. The main contributions are threefold: (1) the correlation between image paths is captured from four views (interaction, distance, motion direction, and motion transition), (2) a multiview clustering method with diversity regularization is proposed to perceive the complementary information within the multiview data and alleviate the influence of redundant features, and (3) a cluster merging strategy is designed to combine the highly correlated clusters and determine the final groups automatically. Experimental results on several benchmark datasets validate the good performance of the proposed method.

Crowding, attention and consciousness: In support of the inference hypothesis

One of the most important topics in current work on consciousness is what relationship it has to attention. Recently, one of the focuses of this debate has been on the phenomenon of identity crowding. Ned Block has claimed that identity crowding involves conscious perception of an object that we are unable to pay attention to. In this article, we draw upon a range of empirical findings to argue against Block's interpretation of the data. We also argue that current empirical evidence strongly supports an interpretation of the data that emphasises cognitive inference over conscious perception.

The madness of crowds phenomenon in the collective decision-making by the cells, the cell’s metacognition and cancer

Recently it has been found that the collective decision-making in the group is efficient only when the confidences (a version of metacognition) of the members are similar, and it has been assumed that the metacognition (self-reference) in general is crucial for the human cooperation. Our goal is to map the decision making by the cells to decision making by humans, looking the analog of metacognition in the cells, accurately calculate the collective sensing of chemical gradients by the cells, and apply our results to cancer. We formulated the model for the chemeosensing by the cells with different diameters, solved it accurately and found that the collective chemosensing is very similar to the collective decision making by humans. We found that the collective sensing of the ligand concentration can be worse than for the most sensitive cell. We introduced the metacognition of the cells, and verify that the metacognition is impaired for the cancer case. We assume as a hypothesis that the impaired cell metacognition in case of cancer does not allow normal multi-cellularity, and cancer can arise when the ”two heads are better than one” principle fails, and there is a ”madness of crowds” phenomenon instead.

Automated Solutions for Crowd Size Estimation

The crowd phenomenon frequently occurs in dense urban living environments. Crowd counting or estimation helps to develop management strategies such as designing safe public places and evacuation plan for emergencies. These strategies are different depending upon the type of event such as political and public demonstrations, sports, and religious events. However, estimating the number of people in crowds at closed or open environments is quite challenging because of the dynamics involved in the process. In addition, crowd estimation itself poses challenges due to randomness in crowd behavior, motion, and an area’s geometric specifications. Crowd behavior as well as the area parameters is studied before suggesting any possible technological solution for managing a crowd. This article presents a theoretical understanding of the major crowd size estimation approaches that cannot be achieved through the study of existing survey papers in this area, because the existing survey papers focus on particular technologies/specific areas with no or brief description of the involved steps. Besides, this article also highlights the strength and weakness of crowd size estimation solutions and their possible applications. It is, therefore, believed that the provided information would assist in developing an intelligent system for crowd management.

Emerging patterns in crowd streams and the aid of ABM for egress management

The aim of this study is to illustrate the advantage of using agent based models (ABM) in order to deepen the understanding of crowd phenomena and help manage design strategies and egress operations on the field. The crowd stream, recognised as a proper nonlinear system, is in fact able to produce characteristic patterns that interact with the environment in recurring shapes. In particular, the free stream and the multi-leader stream are chosen as two mostly representative patterns that develop under typical conditions in a real life egress environment, so that the ABM engineering may bring an added value also in the development of the environment layout. The first part of the study describes the features of these emerging patterns in one geometrical enclosure typical of a Mall. Such issues as queuing, clogging generation, paths of crowd spreading are, for instance, the primary goal upon which to compare the modelling. The overall outcomes, in terms of travel times and people fluxes, complete the scene. The second part of the study describes the connected evolution of these emerging patterns in an environment typical of the landside space in an Airport terminal. Such issues as people advancement, emotions transport, and change of direction are, for instance, the primary goal for the modelling. The overall outcomes of the travel times and people fluxes are the remaining issues to be assessed. The comparison between the results coming from the various ABM packages (Evac, Pathfinder) completes the study, showing how a proper choice in modelling can be based depending on the foreseen goal of the design stage.

How cognitive heuristics can explain social interactions in spatial movement.

The movement of pedestrian crowds is a paradigmatic example of collective motion. The precise nature of individual-level behaviours underlying crowd movements has been subject to a lively debate. Here, we propose that pedestrians follow simple heuristics rooted in cognitive psychology, such as 'stop if another step would lead to a collision' or 'follow the person in front'. In other words, our paradigm explicitly models individual-level behaviour as a series of discrete decisions. We show that our cognitive heuristics produce realistic emergent crowd phenomena, such as lane formation and queuing behaviour. Based on our results, we suggest that pedestrians follow different cognitive heuristics that are selected depending on the context. This differs from the widely used approach of capturing changes in behaviour via model parameters and leads to testable hypotheses on changes in crowd behaviour for different motivation levels. For example, we expect that rushed individuals more often evade to the side and thus display distinct emergent queue formations in front of a bottleneck. Our heuristics can be ranked according to the cognitive effort that is required to follow them. Therefore, our model establishes a direct link between behavioural responses and cognitive effort and thus facilitates a novel perspective on collective behaviour.

A sociologist looks at crowds: Innovation or invention?

Crowd-based organizational models are purported to be more open and participatory than traditional organizational forms. But are they novel inventions or permutations of forms that have existed previously? This essay examines the wide array of innovations pursued under the umbrella label of crowd phenomena and asks whether they have altered traditional ways of organizing. The ramifications of crowds for both workers and consumers are also discussed. Central features of crowd organizing include spot transactions, short-term relations, demand-based pricing, heterogeneous demand, and reputations established through feedback mechanisms. Security and formality appear to have been replaced by openness and precariousness. The essay concludes with a call for further study of the contents of crowd-generated products and services.

Phenomenological description of deadlock formation in pedestrian bidirectional flow based on empirical observation

.This article presents an empirical investigation for the phenomenon of deadlock formation by analyzing a pedestrian bidirectional flow of a subway station in central Tokyo. Using video images obtained from cameras placed in the station during morning rush hour, the in- and outflow (making up the total flow) resulting in a test section considered were computed. With the use of additional information such as train arrival times, test section density and its flow ratio, the formation of different complex crowd phenomena is analyzed. The information provided by the unsteady flow curves allowed a qualitative and quantitative distinction between free flow, congested flow and deadlock formation. This latter phenomenon could be easily identified based on the shape of the total outflow showing a double peak. Based on the information gained, a flow regime diagram using flow and counter-flow values to distinguish between different flow regimes in bidirectional flow has been derived. In this regard, flow ratio was confirmed as being a relevant parameter to describe bidirectional flow, providing experimental evidence that total flow alone can be misleading. Qualitative and quantitative results obtained in the present study were finally compared with previous literature showing good agreement.

Macro-level information transfer in social media: Reflections of crowd phenomena

Online social interactions have become more dynamic and far-reaching across multiple social media platforms than ever before. This is because of frequently changing online contact networks and increasing accessibility to diverse information sources outside of the networks. Accordingly, massive user-generated content spreads through heterogeneous social networks beyond a single social platform. The main goal of this paper is to propose a model-free approach for estimating macro-level information transfer across heterogeneous populations without any assumptions on such dynamic and complex social networks. With this approach, we estimate macro-level diffusion across mainstream news (News), social networking sites (SNS), and blogs (Blog), and the estimations are compared with outcomes from our previous model-driven approach. We also analyze crowd phenomena in diffusion for News, SNS, and Blog as online social systems in terms of activity, reactivity, and heterogeneity. We find that News is the most active, SNS is the most reactive, and Blog is the most persistent, which governs time-evolving heterogeneity. Discovered crowd phenomena are interpreted with respect to our proposed approaches. The strength and directionality of influence reflect reactivity, while topic-related diffusion patterns reflect heterogeneity. We expect that this study can provide a consistent way of understanding cross-population diffusion in diverse application domains.

Current Crowding Phenomenon: Theoretical and Direct Correlation With the Efficiency Droop of Light Emitting Diodes by a Modified ABC Model

By incorporating current crowding effect into the conventional ABC model, we theoretically and directly uncover the black box of the relationship between the current diffusion and the internal quantum efficiency, light extraction efficiency, and external quantum efficiency droop of the light emitting diodes (LEDs). A general formula for the current diffusion length ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$L_{s}$ </tex-math></inline-formula> ) has been derived. The design strategies on the current blocking layer and the transparent conductive layer to increase the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$L_{s}$ </tex-math></inline-formula> have been analyzed. The impact of the vertical resistance ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$R_{s}$ </tex-math></inline-formula> ) on the LED wall-plug efficiency (WPE) has also been addressed. We have found there exists an optimized <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$R_{s}$ </tex-math></inline-formula> , and the LED WPE is not always monotonically increasing as the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$R_{s}$ </tex-math></inline-formula> decreases. Besides, our previous experiment results have been summarized to verify the theoretical analysis.

Crowded Scene Analysis: A Survey

Automated scene analysis has been a topic of great interest in computer vision and cognitive science. Recently, with the growth of crowd phenomena in the real world, crowded scene analysis has attracted much attention. However, the visual occlusions and ambiguities in crowded scenes, as well as the complex behaviors and scene semantics, make the analysis a challenging task. In the past few years, an increasing number of works on crowded scene analysis have been reported, covering different aspects including crowd motion pattern learning, crowd behavior and activity analysis, and anomaly detection in crowds. This paper surveys the state-of-the-art techniques on this topic. We first provide the background knowledge and the available features related to crowded scenes. Then, existing models, popular algorithms, evaluation protocols, as well as system performance are provided corresponding to different aspects of crowded scene analysis. We also outline the available datasets for performance evaluation. Finally, some research problems and promising future directions are presented with discussions.

Universal power law governing pedestrian interactions.

Human crowds often bear a striking resemblance to interacting particle systems, and this has prompted many researchers to describe pedestrian dynamics in terms of interaction forces and potential energies. The correct quantitative form of this interaction, however, has remained an open question. Here, we introduce a novel statistical-mechanical approach to directly measure the interaction energy between pedestrians. This analysis, when applied to a large collection of human motion data, reveals a simple power-law interaction that is based not on the physical separation between pedestrians but on their projected time to a potential future collision, and is therefore fundamentally anticipatory in nature. Remarkably, this simple law is able to describe human interactions across a wide variety of situations, speeds, and densities. We further show, through simulations, that the interaction law we identify is sufficient to reproduce many known crowd phenomena.

The hierarchical sparse selection model of visual crowding.

Because the environment is cluttered, objects rarely appear in isolation. The visual system must therefore attentionally select behaviorally relevant objects from among many irrelevant ones. A limit on our ability to select individual objects is revealed by the phenomenon of visual crowding: an object seen in the periphery, easily recognized in isolation, can become impossible to identify when surrounded by other, similar objects. The neural basis of crowding is hotly debated: while prevailing theories hold that crowded information is irrecoverable – destroyed due to over-integration in early stage visual processing – recent evidence demonstrates otherwise. Crowding can occur between high-level, configural object representations, and crowded objects can contribute with high precision to judgments about the “gist” of a group of objects, even when they are individually unrecognizable. While existing models can account for the basic diagnostic criteria of crowding (e.g., specific critical spacing, spatial anisotropies, and temporal tuning), no present model explains how crowding can operate simultaneously at multiple levels in the visual processing hierarchy, including at the level of whole objects. Here, we present a new model of visual crowding—the hierarchical sparse selection (HSS) model, which accounts for object-level crowding, as well as a number of puzzling findings in the recent literature. Counter to existing theories, we posit that crowding occurs not due to degraded visual representations in the brain, but due to impoverished sampling of visual representations for the sake of perception. The HSS model unifies findings from a disparate array of visual crowding studies and makes testable predictions about how information in crowded scenes can be accessed.

Sharp interfaces in two-dimensional free boundary problems: interface calculation via matched conformal maps.

We use conformal maps to study a free boundary problem for a two-fluid electromechanical system, where the interface between the fluids is determined by the combined effects of electrostatic forces, gravity, and surface tension. The free boundary in our system develops sharp corners or singularities in certain parameter regimes, and this is an impediment to using existing "single-scale" numerical conformal mapping methods. The difficulty is due to the phenomenon of crowding, i.e., the tendency of nodes in the preimage plane to concentrate near the sharp regions of the boundary, leaving the smooth regions of the boundary poorly resolved. A natural idea is to exploit the scale separation between the sharp regions and smooth regions to solve for each region separately and then stitch the solutions together. However, this is not straightforward as conformal maps are rigid "global" objects, and it is not obvious how one would patch two conformal maps together to obtain a new conformal map. We develop a "multiscale" (i.e., adaptive) conformal mapping method that allows us to carry out this program of stitching conformal maps on different scales together. We successfully apply our method to the electromechanical model problem.