Dynamically Measuring, Analyzing and Forecasting Sentiment Resilience in Online Communities

About Our Authors

One of the major hurdles preventing the full exploitation of information from online communities is the widespread concern regarding the quality and credibility of user-contributed content. Prior works in this domain operate on a static snapshot of the community, making strong assumptions about the structure of the data (e.g., relational tables), or consider only shallow features for text classification. To address the above limitations, we propose probabilistic graphical models that can leverage the joint interplay between multiple factors in online communities --- like user interactions, community dynamics, and textual content --- to automatically assess the credibility of user-contributed online content, and the expertise of users and their evolution with user-interpretable explanation. To this end, we devise new models based on Conditional Random Fields for different settings like incorporating partial expert knowledge for semi-supervised learning, and handling discrete labels as well as numeric ratings for fine-grained analysis. This enables applications such as extracting reliable side-effects of drugs from user-contributed posts in healthforums, and identifying credible content in news communities. Online communities are dynamic, as users join and leave, adapt to evolving trends, and mature over time. To capture this dynamics, we propose generative models based on Hidden Markov Model, Latent Dirichlet Allocation, and Brownian Motion to trace the continuous evolution of user expertise and their language model over time. This allows us to identify expert users and credible content jointly over time, improving state-of-the-art recommender systems by explicitly considering the maturity of users. This also enables applications such as identifying helpful product reviews, and detecting fake and anomalous reviews with limited information.

Purpose– The purpose of this paper is to explore the dynamics of an online community by examining its participants’ centrality measures: degree, closeness, and the betweenness centrality. Each centrality measure shows the different roles and positions of an individual participant within an online community. To be specific, this research examines how an individual participant’s role and position affects her/his information sharing activities within an online community over time. Additionally, it investigates the differences between two different online communities (a personal interest focussed community and a social interest focussed community), in terms of the interaction patterns of participants.Design/methodology/approach– For this research, the authors collected log files from Korean online discussion communities (café.naver.com) using a crawler program. A social network analysis was used to explore the interaction patterns of participants and calculate the centrality measures of individual participants. Time series cross-sectional analysis was used to analyze the effects of the roles and the positions on their information sharing activities in a longitudinal setting.Findings– The results of this research showed that all three centrality measures of an individual participant in previous time periods positively influenced his/her information sharing activity in the current periods. In addition, this research found that, depending on the nature of the discussion issues, the participants showed different interaction patterns. Throughout this research, the authors explored the interaction patterns of individual participants by using a network variable, the centrality, within a large online community, and found that the interaction patterns provided strong impact on their information sharing activities in the following months.Research limitations/implications– To investigate the changes of participant’s behaviors, this study simply relies on the numbers of comments received and posted without considering the contents of the comments. Future studies might need to analyze the contents of the comments exchanged between participants, as well as the social network among participants.Practical implications– Online communities have developed to take a more active role in inviting public opinions and promoting discussion about various socio-economic issues. Governments and companies need to understand the dynamics which are created by the interactions among many participants. This study offers them a framework for analyzing the dynamics of large online communities. Furthermore, it helps them to respond to online communities in the right way and in the right time.Social implications– Online communities do not merely function as a platform for the free exchange and sharing of personal information and knowledge, but also as a social network that exerts massive influence in various parts of society including politics, economy, and culture. Now online communities become playing an important role in our society. By examining communication or interaction behaviors of individual participants, this study tries to understand how the online communities are evolved over time.Originality/value– In the area of online communities, many previous studies have relied on the subjective data, like participant’s perception data, in a particular time by using survey or interview. However, this study explores the dynamics of online communities by analyzing the vast amount of data accumulated in online communities.

Catastrophe theory and patterns in 30-second freeway traffic data— Implications for incident detection

Catastrophe theory is a new field in mathematical topology that allows the formulation of comprehensive qualitative models of systems which have previously eluded rigorous mathematical formulation. Because the models have a topological foundation, many seem ingly dissimilar phenomena can be related to a common underlying topological structure. The properties of that structure can then be studied in a convenient form and the conclusions related back to the original problem. This paper provides an introduction to catastrophe theory and defines the principal condi tions required for its application. The basic prop erties of bimodality, discontinuity (catastrophe), hysteresis, and divergence are defined and illustra ted using the simplest structures of the theory. In this paper the spruce budworm/forest ecosystem of eastern Canada illustrates how catastrophe theory may be applied in ecology. With only a minimum of general descriptive information about the budworm system, a qualitative catastrophe theory model is hypothesized. This model is rich in its ability to provide predictions and insight about the global behavior of the system. To further check and verify the assumptions of this qualitative model, an exist ing detailed simulation model is analyzed from the perspective of catastrophe theory and is found to exhibit a basic underlying structure that is very similar to the catastrophe theory model. A detailed examination of the simulation model shows that catas trophe theory provides a consistent framework for analyzing and interpreting the behavior of this eco system.

In this paper, the base-band signal collected from an unknown, multipath, multi-receiver FIR channel is viewed as a state sequence generated by a hidden Markov model (HMM) whose states and order are unknown and whose transition probability matrix with an unknown permutation is known once the order is given. Based on this view, two types of algorithms are developed for acquisition and tracking, respectively. The algorithms are suitable for both block and non-block transmissions, and for time-varying channels. For acquisition the states and the transition probability matrix of a fully-connected HMM with a possible lower order are estimated by using a clustering algorithm. Then a state sequence is estimated based on a maximum a posterior (MAP) estimator using the Viterbi algorithm with a fully-connected trellis. This state sequence is used to refine the estimated states and transition probability matrix of the HMM. Based on the fully-connected HMM, the states are properly assigned to symbol vectors and the non-fully-connected HMM is determined, which is used in the maximum likelihood (ML) sequence estimation. In tracking, the symbol sequence is estimated by the Viterbi algorithm, and the states of the HMM are updated at each release of estimated states. Simulation results show that the proposed blind algorithms with the lower-ordered HMM achieve a performance comparable with the ML estimator utilizing the known, channel parameters. Theoretical analysis confirms simulation results.

Wear fault is one of the failure modes most frequently occurring in gears. Identifying different wear levels, especially for early wear is a challenge in gear fault diagnosis. In this paper, a machine learning methodology for performing diagnosis is presented. It is developed based on hidden Markov model (HMM). A HMM is defined as a doubly stochastic process. The underlying stochastic process is a discrete time finite-state homogeneous Markov chain. The state sequence is not observable and so is called hidden. HMM's construction process and the condition monitoring process are same. So, the HMM can be used to model the mechanical fault progress process. There are some authors applying HMM in machine fault diagnosis. However, HMM with different observation styles applying in machine fault diagnosis will produce different results. With enough history data, this paper aims to using these different HMMs to classify the different levels of gear wears automatically and analysis the performance of these models. These observation styles are discrete integer observation which is the quantized fault features extracted from the vibration signals, continuous observation which is the feature with no quantizing extracted from the vibration signals and observation which is the preprocessed vibration signals. The gear wear experiments were conducted and the vibration signals were captured from the gears under different loads and motor speeds. HMM with three different styles observation is applied to identify the gear wears and the applied results are analyzed. The results show that the features with no quantizing as the observations of the HMM for diagnosis perfected best. But, at the case of no history data, HMM which the observations are the preprocessed vibration signals is used to do the incipient fault detection. The result shows the validity of the method.

Introduction Most developed countries, including Australia, have a strong focus on national, state and local strategies for emergency management and response in the face of disasters and crises. This framework can include coping with catastrophic dislocation, service disruption, injury or loss of life in the face of natural disasters such as major fires, floods, earthquakes or other large-impact natural events, as well as dealing with similar catastrophes resulting from human actions such as bombs, biological agents, cyber-attacks targeting essential services such as communications networks, or other crises affecting large populations. Emergency management frameworks for crisis and disaster response are distinguished by their focus on the domestic context for such events; that is, how to manage and assist the ways in which civilian populations, who are for the most part inexperienced and untrained in dealing with crises and disasters, are able to respond and behave in such situations so as to minimise the impacts of a catastrophic event. Even in countries like Australia that demonstrate a strong public commitment to cultural pluralism and social cohesion, ethno-cultural diversity can be seen as a risk or threat to national security and values at times of political, natural, economic and/or social tensions and crises. Australian government policymakers have recently focused, with increasing intensity, on “community resilience” as a key element in countering extremism and enhancing emergency preparedness and response. In some sense, this is the result of a tacit acknowledgement by government agencies that there are limits to what they can do for domestic communities should such a catastrophic event occur, and accordingly, the focus in recent times has shifted to how governments can best help people to help themselves in such situations, a key element of the contemporary “resilience” approach. Yet despite the robustly multicultural nature of Australian society, explicit engagement with Australia’s cultural diversity flickers only fleetingly on this agenda, which continues to pursue approaches to community resilience in the absence of understandings about how these terms and formations may themselves need to be diversified to maximise engagement by all citizens in a multicultural polity. There have been some recent efforts in Australia to move in this direction, for example the Australian Emergency Management Institute (AEMI)’s recent suite of projects with culturally and linguistically diverse (CALD) communities (2006-2010) and the current Australia-New Zealand Counter-Terrorism Committee-supported project on “Harnessing Resilience Capital in Culturally Diverse Communities to Counter Violent Extremism” (Grossman and Tahiri), which I discuss in a longer forthcoming version of this essay (Grossman). Yet the understanding of ethno-cultural identity and difference that underlies much policy thinking on resilience remains problematic for the way in which it invests in a view of the cultural dimensions of community resilience as relic rather than resource – valorising the preservation of and respect for cultural norms and traditions, but silent on what different ethno-cultural communities might contribute toward expanded definitions of both “community” and “resilience” by virtue of the transformative potential and existing cultural capital they bring with them into new national and also translocal settings. For example, a primary conclusion of the joint program between AEMI and the Australian Multicultural Commission is that CALD communities are largely “vulnerable” in the context of disasters and emergency management and need to be better integrated into majority-culture models of theorising and embedding community resilience. This focus on stronger national integration and the “vulnerability” of culturally diverse ethno-cultural communities in the Australian context echoes the work of scholars beyond Australia such as McGhee, Mouritsen (Reflections, Citizenship) and Joppke. They argue that the “civic turn” in debates around resurgent contemporary nationalism and multicultural immigration policies privileges civic integration over genuine two-way multiculturalism. This approach sidesteps the transculturational (Ortiz; Welsch; Mignolo; Bennesaieh; Robins; Stein) aspects of contemporary social identities and exchange by paying lip-service to cultural diversity while affirming a neo-liberal construct of civic values and principles as a universalising goal of Western democratic states within a global market economy. It also suggests a superficial tribute to cultural diversity that does not embed diversity comprehensively at the levels of either conceptualising or resourcing different elements of Australian transcultural communities within the generalised framework of “community resilience.” And by emphasising cultural difference as vulnerability rather than as resource or asset, it fails to acknowledge the varieties of resilience capital that many culturally diverse individuals and communities may bring with them when they resettle in new environments, by ignoring the question of what “resilience” actually means to those from culturally diverse communities. In so doing, it also avoids the critical task of incorporating intercultural definitional diversity around the concepts of both “community” and “resilience” used to promote social cohesion and the capacity to recover from disasters and crises. How we might do differently in thinking about the broader challenges for multiculturalism itself as a resilient transnational concept and practice? The Concept of Resilience The meanings of resilience vary by disciplinary perspective. While there is no universally accepted definition of the concept, it is widely acknowledged that resilience refers to the capacity of an individual to do well in spite of exposure to acute trauma or sustained adversity (Liebenberg 219). Originating in the Latin word resilio, meaning ‘to jump back’, there is general consensus that resilience pertains to an individual’s, community’s or system’s ability to adapt to and ‘bounce back’ from a disruptive event (Mohaupt 63, Longstaff et al. 3). Over the past decade there has been a dramatic rise in interest in the clinical, community and family sciences concerning resilience to a broad range of adversities (Weine 62). While debate continues over which discipline can be credited with first employing resilience as a concept, Mohaupt argues that most of the literature on resilience cites social psychology and psychiatry as the origin for the concept beginning in the mid-20th century. The pioneer researchers of what became known as resilience research studied the impact on children living in dysfunctional families. For example, the findings of work by Garmezy, Werner and Smith and Rutter showed that about one third of children in these studies were coping very well despite considerable adversities and traumas. In asking what it was that prevented the children in their research from being negatively influenced by their home environments, such research provided the basis for future research on resilience. Such work was also ground-breaking for identifying the so-called ‘protective factors’ or resources that individuals can operationalise when dealing with adversity. In essence, protective factors are those conditions in the individual that protect them from the risk of dysfunction and enable recovery from trauma. They mitigate the effects of stressors or risk factors, that is, those conditions that predispose one to harm (Hajek 15). Protective factors include the inborn traits or qualities within an individual, those defining an individual’s environment, and also the interaction between the two. Together, these factors give people the strength, skills and motivation to cope in difficult situations and re-establish (a version of) ‘normal’ life (Gunnestad). Identifying protective factors is important in terms of understanding the particular resources a given sociocultural group has at its disposal, but it is also vital to consider the interconnections between various protective mechanisms, how they might influence each other, and to what degree. An individual, for instance, might display resilience or adaptive functioning in a particular domain (e.g. emotional functioning) but experience significant deficits in another (e.g. academic achievement) (Hunter 2). It is also essential to scrutinise how the interaction between protective factors and risk factors creates patterns of resilience. Finally, a comprehensive understanding of the interrelated nature of protective mechanisms and risk factors is imperative for designing effective interventions and tailored preventive strategies (Weine 65). In short, contemporary thinking about resilience suggests it is neither entirely personal nor strictly social, but an interactive and iterative combination of the two. It is a quality of the environment as much as the individual. For Ungar, resilience is the complex entanglements between “individuals and their social ecologies [that] will determine the degree of positive outcomes experienced” (3). Thinking about resilience as context-dependent is important because research that is too trait-based or actor-centred risks ignoring any structural or institutional forces. A more ecological interpretation of resilience, one that takes into a person’s context and environment into account, is vital in order to avoid blaming the victim for any hardships they face, or relieving state and institutional structures from their responsibilities in addressing social adversity, which can “emphasise self-help in line with a neo-conservative agenda instead of stimulating state responsibility” (Mohaupt 67). Nevertheless, Ungar posits that a coherent definition of resilience has yet to be developed that adequately ‘captures the dual focus of the individual and the individual’s

Creativity in an Online Community as a Response to the Chaos of a Breast Cancer Diagnosis

The purpose of this research is to consider the applicability of catastrophe theory to research in higher education. The review of the literature concentrates on those efforts in which the variables and their use have an analogy to higher education. Several problems which typically appear in the literature are presented in a theoretical framework, and a catastrophe theory model is attempted for each. In this way the researchers were able to consider the fundamental assumptions about a given problem and its variables and to assess, at a theoretical level, whether catastrophe theory might be an appropriate tool for analysis.

Discontinuities as a crucial aspect of economic systems have been discussed both verbally—particularly in institutionality theory—and formally, chiefly using catastrophe theory. Catastrophe theory has, however, been criticized heavily for lacking micro-foundations and has mainly fallen out of use in economics and social sciences. The present paper proposes a simple catastrophe theory model of technological change with network externalities and reevaluates the value of such a model by adding an agent-based micro layer. To this end an agent-based variant of the model is proposed and investigated specifically with regard to the network structure among the agents. While the macro level of the model produces a classical cusp catastrophe—a result that is preserved in the agent-based form—it is found that the behavior of the model changes locally depending on the network structure, especially if networks with features that resemble social networks (low diameter, high clustering, power law distributed node degree) are considered. While the present work investigates merely an aspect out of a large possibility space, it encourages further research using agent-based catastrophe theory models especially of economic aspects to which catastrophe theory has previously successfully been applied; aspects such as technological and institutional change, economic crises, or industry structure.

With the availability of the Internet, virtual communities are proliferating at an unprecedented rate. In-depth understanding of virtual community dynamics can help us to address critical organizational and information systems issues such as communities-of-practice, virtual collaboration, and knowledge management. The biggest challenge in fostering a virtual community is the supply of knowledge through services and the willingness to share knowledge with other members. This paper integrates the Service Oriented Architecture and Agent Based Theory to construct a Reinforced Concrete design (RCD) model for Reinforced Concrete Analysts and Designers in virtual communities. The aim of this paper is to elucidate the sharing of knowledge in virtual community from the perspective of RCD serviceability service. Results confirm that RC Designers will share knowledge; if their perceive benefit exceed the cost of their sharing behaviour. This study is useful for the developers of Service oriented Applications for virtual communities to insight into knowledge sharing in cyberspace for RCD.

The participation of individual users in online communities is one of the most noted features in the recent explosive growth of popular online communities ranging from picture and video sharing (Flickr.com and YouTube.com) and collective music recommendation (Last.fm) to news voting (Digg. com) and social bookmarking (del.icio.us). Unlike traditional online communities, these sites feature little message exchange among users. Nevertheless, users' involvement and their contribution through non-message-based interactions have become a major force behind successful online communities. Recognition of this new type of user participation is crucial to understanding the dynamics of online social communities and community monetization. The new communication features in online communities can be best summarized as Ballot Box Communication (BBC), which is an aggregation mechanism that reflects the common experience and opinions among individuals. By offering a limited number of choices such as voting, rating and tagging, BBC creates a new medium to effectively reveal the interests of mass population (see Table 1). Compared with traditional Computer Mediated Communication (CMC) such as email, Web publishing, and online forums, BBC influences user preferences by simplifying the mass sharing of individual preferences. These technologies offer new ways for information consumers to be involved in community activities. In traditional online communities, users only have two levels of participation: "watching from the sidelines" or "playing in the game," for example, they are either passive readers or active participants in conversations. However, BBC presents a new choice -- "shouting from the stands" -- in which each user can express his opinion through BBC and their collective preferences can be heard as a dominant voice. For instance, Digg readers can vote on news and promote it to the front page for millions of visitors to see. In spite of the increasing significance of non-message-based online communication, very little is known about BBC-enabled communities. As entrepreneurs build and manage new online communities, they have no choice but to look for the "right" technologies by trial-and-error. Not surprisingly, the result is hit-or-miss: some of the grandest failures of the dot com bust featured online communities. Only after costly failures, it has been recognized that not all technologies can benefit the growth and sustainability of a community. Extant theories on online communities and communication networks may offer some guidance on understanding of the emergence of new online communities (such as YouTube). Whitaker et al. identify online communities as "intense interactions, strong emotional ties and shared activities." In addition, Monge and Contractor define communication networks as "the patterns of contact that are created by the flow of messages among communicators through time and space." Both study the social interaction aspect of communities such as user commenting and discussing. However, the nonsocial interaction aspect, which is the focus of BBC and often dominant in contemporary online communities, has not received much attention.

This paper describes an automatic text‐ and speaker‐independent language identification method based on hidden Markov models (HMMs) for acoustic features. The hidden Markov modeling is used to represent the phonotactics for each language. Each language has its proper phonotactics.The HMM topology here is a fully structured (ergodic) model wherein any state could transit to all states. Two kinds of HMMs are used: the discrete HMM (DHMM) with the codebook and the continuous density HMM (CHMM). The HMM was trained using both the Baum‐Welch (forward‐backward) algorithm and the Viterbi algorithm. The latter was used for emphasizing the state transition probability.For comparison, experiments also were conducted on the identification using a mixtured Gaussian distribution model with one state. This single‐state Gaussian distribution model gave the same performance as the HMM trained with the Baum‐Welch algorithm. This is because the transition between states which reflects the characteristics of each language does not affect the likelihood scores very much. This problem was addressed by emphasizing the transition probabilities and using the Viterbi algorithm, which resulted in an improvement in the recognition rates. The trigram for optimal state sequence is introduced. Combining it with the HMM produced the best results.

Hidden Markov models (HMMs) are widely used by many applications for forecasting purposes. They are increasingly becoming popular models as part of prediction systems in finance, marketing, bio-informatics, speech recognition, signal processing, and so on. Given an HMM, an application of HMMs is to choose a state sequence so that the joint probability of an observation sequence and a state sequence given the model is maximized. Although this seems an easy task if the model is given, it becomes a challenge when the model is distributed between various parties. Due to privacy, financial, and legal reasons, the model owners might not want to integrate their split models. In this paper, we propose schemes to select a state sequence so that the joint probability of an observation sequence and a state sequence given the model is maximized when the model is horizontally or vertically distributed between two parties while preserving their privacy. We then analyze the proposed schemes in terms of privacy, accuracy, and additional overhead costs. Since privacy, accuracy, and performance are conflicting goals, our proposed methods are able to achieve an equilibrium among them.