A Twitter Social Media Modeling: Tesla’s First Fatal Crash Tweeting Behavior

In recent years, new social network data has been available due to adoption of new Information and Communication Technologies (ICT) that have allowed the development of social media networks. Twitter, as one of the most popular social networks around the globe, offers particularly large public data sets available by Twitter's RESTful Application Programming Interface (API) that allow developers to request through HTTP access to core Twitter data, including update timelines, status data, and user information. Through the development of a program script that specially run-time environment that automates the retrieve of Twitter information was possible to store and model a heterogeneous-multiplex network in Neo4j graph-databaseengine. The objective of this project is to be able to describe and explore the social media network topology from a specific business application. This was done by using Python and Neo4j modeling of social media interactions with the objective of understand social and behavioral relationships by social network analysis

AUTHOR: J.A. Sánchez Castro

Competitive Networks

Creating a visualization tool, we propose an approach to find patterns in competitive networks. The case study was of data on soccer games obtained from FIFA 2010 World Cup. The information was used to model networks of winning and losing teams based on different measures to show their performance. This visualization tool gives an alternative perspective to find characteristics of the playing style and organization of winning teams. This tool can also help to predict winners and losers for a match if patterns are repeated.

AUTHOR: Denisse Martinez Mejorado

Framework to Assess and Visualize the Likelihood of Sewer System Failure

To overcome accidental breakages and sewer collapses due to the aging of sewer systems, quantitative approaches are used to identify the assets with the highest likelihood of failure. Because of the lack of availability of extensive datasets in practice for most authorities, qualitative models can be sufficient to strategically plan renovation.

This paper presents a qualitative method to identify weak spots within a sewer system and represent the results in a visual manner. It combines multiple information and influence factors to create heat maps of sewer failure to improve the effectiveness of sewer pipe replacement. The method was applied to a sewer system located in Hoboken (NJ, USA) upon calibration with local expert knowledge and validated with Closed Circuit Television (CCTV) images. The model agrees well with the data, opening the possibility to be applied to other sewer systems for similar studies.

Keywords: aging infrastructure; data visualization; decision making; sewer asset management.

Authors: M.J.W. Bach & N.B. Hoving

 

Crowdsourcing Clustering

Clustering in datasets usually has been made by algorithms that found similarity measures in the data, but what if users can have an interface where they can select the clusters and give the work to people, crowdsource their clusters and compare it with common algorithms.

Keywords- Clustering, Crowdsourcing, Multivariate analysis, Parallel Coordinates, Andrews Plot.

AUTHOR: Samuel Rocha

Simulation of Evacuation in Case of Natural Disasters

Evacuation planning is fundamental to ensure that most people can be evacuated to a safe area when a natural disaster occurs. This paper takes the city of Hoboken. The main challenge in this paper is to determine which are the main variables that a ffects the routing to safe areas. The paper shows the evacuation network within a city model described in Aimsun Simulator. We propose an Aimsun scenarios described in microscopic scenarios under di fferent restrictions that approach to solve the evacuation routing through Hoboken City. This problem involves objectives that need to be achieved simultaneously, such as time minimization of total evacuation time, minimal cumulative traffic congestion, which are the main factors that allow an evacuation to be more time efficient and How does the evacuation affect the population's resilience or ability to recover.

AUTHOR: Edwin Martin Cruz Colin

Mapping the Decision-Making Process of Post-Disaster Resilience Projects

This project explores some of the determining features of effective and rapid decision-making processes in post-disaster resilience projects. It aims to examine how established theory on public decision-making can be applied in two post-disaster resilience projects and whether this theory is relevant in varying contexts based on the logic of realist explanation. The cases analyzed in this study are the post-disaster resilience projects in the neighborhood Roombeek in Enschede (Netherlands) after it was largely destroyed by the explosion of a firework depot and the Hudson River Project in Hoboken, New Jersey (USA) after Hurricane Sandy led to severe flooding of the city. The lessons about public decision-making processes that can be learned from studying Roombeek and Hoboken can provide beneficial insights for the future of the Hudson River project and other similar projects. Mechanisms observed in both contexts might also apply to similar projects in the future and help public administrators to increase the effectiveness and efficiency of decision making processes.

AUTHORS: Oliver Klinkhammer, Mathias Quickert, Marie Helen Ferdelman