Self-Exciting Point Process Modeling of Conversation Event Sequences

Masuda, Naoki; Takaguchi, Taro; Sato, N.; Yano, Kazuo

doi:10.1007/978-3-642-36461-7_12

Cited by 43 publications

(42 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The applicability of Hawkes process for modeling and analyzing time-series data has stimulated interest in diverse disciplines, including seismology (Ogata, 1988(Ogata, , 1998, criminology (Mohler et al, 2011;Egesdal et al, 2010;, finance (Chehrazi and Weber, 2011;Aït-Sahalia et al, 2010;Bacry et al, 2013;Filimonov and Sornette, 2012;Embrechts et al, 2011), computational neuroscience (Johnson, 1996;Krumin et al, 2010), genome sequencing (Reynaud-Bouret and Schbath, 2010) and social networks (Crane and Sornette, 2008;Mitchell and Cates, 2009;Simma and Jordan, 2010;Masuda et al, 2012).…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Reactive point processes: A new approach to predicting power failures in underground electrical systems

Ertekin¹,

Rudin²,

McCormick³

2015

Ann. Appl. Stat.

View full text Add to dashboard Cite

Reactive point processes (RPP's) are a new statistical model designed for predicting discrete events, incorporating self-exciting, self-regulating, and saturating components.The self-excitement occurs as a result of a past event, which causes a temporary rise in vulnerability to future events. The self-regulation occurs as a result of an external "inspection" which temporarily lowers vulnerability to future events. RPP's can saturate when too many events or inspections occur close together, which ensures that the probability of an event stays within a realistic range. RPP's were developed to handle an important problem within the domain of electrical grid reliability: short term prediction of electrical grid failures ("manhole events"), including outages, fires, explosions, and smoking manholes, which can cause threats to public safety and reliability of electrical service in cities. For the self-exciting, self-regulating, and saturating elements of the model, we develop both a nonparametric estimation strategy and introduce a class of flexible parametric functions reflecting how the influence of past events and inspections on vulnerability levels gradually fades over time. We use the model to predict power grid failures in Manhattan over a short term horizon.

show abstract

Section: Related Workmentioning

confidence: 99%

“…A similar model with a memory kernel has been used to model conversation event sequences (Masuda et al, 2012).…”

Section: Related Workmentioning

confidence: 99%

Reactive point processes: A new approach to predicting power failures in underground electrical systems

Ertekin¹,

Rudin²,

McCormick³

2015

Ann. Appl. Stat.

View full text Add to dashboard Cite

show abstract

“…Our model is based on the Hawkes process [40], which was first proposed to model seismic patterns [41] and was recently applied to human communication behavior [30,[42][43][44], In the Hawkes model, the activity rate of an individual, denoted by λ(t), is determined by…”

Section: A Generalized Hawkes Model Incorporating Response Behaviormentioning

confidence: 99%

Input-output relationship in social communications characterized by spike train analysis

et al. 2016

Self Cite

View full text Add to dashboard Cite

We study the dynamical properties of human communication through different channels, i.e., short messages, phone calls, and emails, adopting techniques from neuronal spike train analysis in order to characterize the temporal fluctuations of successive inter-event times. We first measure the so-called local variation (LV) of incoming and outgoing event sequences of users, and find that these in-and out-LV values are positively correlated for short messages, and uncorrelated for phone calls and emails. Second, we analyze the response-time distribution after receiving a message to focus on the input-output relationship in each of these channels. We find that the time scales and amplitudes of response are different between the three channels. To understand the impacts of the response-time distribution on the correlations between the LV values, we develop a point process model whose activity rate is modulated by incoming and outgoing events. Numerical simulations of the model indicate that a quick response to incoming events and a refractory effect after outgoing events are key factors to reproduce the positive LV correlations.

show abstract

“…Figure 3 is a histogram of the number of emails exchanged among the remaining pairs, all of them less than 100. Many of the pairs of officers exchanged only a few emails, while a few pairs exchanged a substantial proportion of all emails in the corpus, and a few users (13,18,22) appear three times or more in this list of highly active pairs. These observations are consistent with a coreperiphery structure, which is a characteristic of many social networks [7].…”

Section: The Ikenet Data Setmentioning

confidence: 99%

“…We take a maximum-likelihood approach, using an expectation-maximisation numerical method to combat the problem's ill conditioning [36]. Finally, we give several examples for different choices of the triggering function g. It is most common in the literature to assume an exponential form for g [5,11,15,22,34], though other forms are also in use, including power law [6,27] and the exponential multiplied by a polynomial [28]. Our comparison of exponential and power-law forms suggests that it does not matter which is used, validating the frequent use of the exponential form.…”

Section: Em Estimation Of Hawkes Process Parametersmentioning

confidence: 99%

Point-process models of social network interactions: Parameter estimation and missing data recovery

et al. 2015

View full text Add to dashboard Cite

Electronic communications, as well as other categories of interactions within social networks, exhibit bursts of activity localised in time. We adopt a self-exciting Hawkes process model for this behaviour. First we investigate parameter estimation of such processes and find that, in the parameter regime we encounter, the choice of triggering function is not as important as getting the correct parameters once a choice is made. Then we present a relaxed maximum likelihood method for filling in missing data in records of communications in social networks. Our optimisation algorithm adapts a recent curvilinear search method to handle inequality constraints and a nonvanishing derivative. Finally we demonstrate the method using a data set composed of email records from a social network based at the United States Military Academy. The method performs differently on this data and data from simulations, but the performance degrades only slightly as more information is removed. The ability to fill in large blocks of missing social network data has implications for security, surveillance, and privacy.

show abstract

Self-Exciting Point Process Modeling of Conversation Event Sequences

Cited by 43 publications

References 43 publications

Reactive point processes: A new approach to predicting power failures in underground electrical systems

Reactive point processes: A new approach to predicting power failures in underground electrical systems

Input-output relationship in social communications characterized by spike train analysis

Point-process models of social network interactions: Parameter estimation and missing data recovery

Contact Info

Product

Resources

About