ISCRAM2016 Highlights Day 1

For those of you interested in any form of disaster management ‘The International Conference for Crisis Response and Management’ is the place for you. This year it’s taking place in Rio de Janeiro, Brazil, on the gorgeous Copacabana beach.

This post will summarise a handful of paper presentations which I was able to attend on the first day of the conference proceedings, and provide you links to these papers.

The day began with a keynote talk from José Holguín-Veras, an academic who has spent his career advising organisations on how best to manage disasters. He discussed the top ten lessons from large disasters and catastrophes, and how these had implications for resiliency and the ISCRAM community.

The message that stuck from this presentation was that common sense should not be overlooked in disaster logistics. Buying and shipping bottled water to a population that had been struck by disaster is the least effective gesture, and instead charities and organisations need to be more realistic: sending the amount of money it cost to buy and ship that water could have bought nearly twice as much from a nearer region.

Following this the structure of the day was then broken up into three main sessions, interspersed with breaks.

monday programme

In the first session I attended a series of talks in the track ‘intelligent decision support in the networked society’ or IDSS. This began with the topic ‘Adoption of big data in crisis management toward a better support in decision-making‘ (Fertier et al., 2016; presented by Frederick Beneben). This took some well-used approaches for big data management and produced and improved methodology for processing, thus resulting in a system that allowed for more informed decisions.

A paper of particular interest to myself was entitled ‘Communication and Tracking Ontology Development for Civilians Disaster Assistance‘(Hassan & Chen-Berger, 2016). In this, the authors develop generalised use of the semantic Web to apply during disasters. Ontologies, it turned out, were very popular across the tracks at the conference this year.

In the later afternoon sessions there were several talks that I enjoyed.

Firstly ‘#geiger 2: Developing Guidelines for Radiation Measurements Sharing on Social Media‘ (Segault et al., 2016; presented by Frederico Tajario). This looked at using Twitter metadata to reduce risk from radiation incidents, which seemed very relevant to a developing world that has seen several nuclear disasters in the past few years such as Fukushima in 2011.

Finally,  ‘An Emotional Step Towards Automated Trust Detection in Crisis Social Media ‘ (Halse et al., 2016; Presented by Andrea Tapia) tackled with some really big issues of the trustworthiness of online information. An original and more sociological approach was taken in this paper, and instead Tweets were analysed and categorised by associated emotion rather than a measure of credibility.

Day 1 was diverse and thoroughly interesting. But, as ever, one cannot attend everything.

 

 

 

 

 

Advertisements

The Human Face of Big Data

I’m imminently about to start my PhD in Web Science, beginning with the MSc. I’m a big ball of nervousness, excitement and the feeling of actually having the energy to do research that I’ll be interested in. One of the bi-products of this however, is that far more people are asking me about my course.

Generally I give them a brief description of how the web has effected things, and will continue to effect things. Often, this confuses people.

I used to get this a lot with my degree in Geography too. People say things like ‘All I remember about geography is Oxbow lakes LOL, what do you actually study? Colouring?’.

This really grated me for a long time, because geography is the study of how everything relates to one another. By the time I had finished my degree, when someone asked what I studied in particular I simply used to repress an eye-twitch and shout ‘Everything! Geography is everything!’ and then skulk off to get a drink.

Consequently I’ve begun to explain things regarding my new MSc and PhD in a way you would justify learning a maths equation – by giving it a context in which it can be applied.

So here are some excellent examples of crossovers between Geography and Web science that I’m interested in researching further. The first one is human-based, the second one physical-based, and the third is a little bit about my other posts:

‘The Human face of Big Data’ is a term used to describe and explain the human nature behind data. It addresses the growing desire to understand what data can tell us about ourselves, and why we use it in the way that we do. It is commonly used in the study of Social media such as Facebook and Twitter, and how human usage can reveal insights in to real-life trends that make up who we collectively are: our cultures, our mindsets, who we favour or disfavour, and much more.

For instance, analysing billions of Tweets helped two researchers record new insights about public health issues and the way disease is spread. Further to this, they were able to tailor specific health warnings, targeting selected sites that the data indicated to them were the sites of choice most individuals would encounter. As a result of the increased connectivity of our culture, metrics and city structures, people were able to spread information regarding health warnings far faster via social media than they would have relying on previous methods (i.e. by watching the news, or by word of mouth).

Additionally the same analysis and smart application of data can be done with physical hazards as well: change out ‘disease’ for a ‘natural hazard’, for example an earthquake, and the same kind of responsiveness of social media can be recorded to similar effect.

The interesting development with using social media as data streams is that one can ask how and why a population may be affected by these notifications. How can the data be used in prediction systems and 3D modelling of communities. It also lends to the development of maps, virtual realities and other visualisations to adapt to different mediums and scholarly discourses.

Producing the most detailed and applicable results is another substantial task, as one is effectively waiting for a natural disaster to occur in order to collect the data. Once this happens the applications it may be used for are endless, and there are certainly other big data sets other than social media that may help to answer, or build on, the same hypotheses.

If you have read any of my other posts you will realise that there is also a pretty cool crossover between geography and the gaming industry. Again, big data is used in game development in order to analyse, assess and tailor games to target audiences.

However, the gaming industry is one of the leading developers of Geographical Virtual Environments (geoVE’s), usage of real geographic data in game play, new methodologies of recreating virtual city-scapes, combined with ever improving graphics.

Great examples of this can be seen nearly continuously over time as games evolve with softwares. Assassins Creed, Skyrim, Civilization, and Grand Theft Auto to name a few are leading not only in real reflective city-scapes, but also in the virtualization of mapping and other geographic data.

The modification of geoVE’s and VR’s (virtual realities) will be a key-stone in the future prediction of disaster management systems. They aim to do this via transforming the visualisation of a predicted event and tailoring responses to it.

By hosting geoVE’s using gaming platforms and online big data storage, a more accurate and intricate representation of city metrics, human nature and city infrastructure can be mapped out in a way never before achieved.

Certainly, having more information from multiple sources which are able to update instantaneously would be a viable and productive goal in real-time disaster management.

Hopefully this has given you more of an idea of the areas I will be studying, and some of the content that is to come on this blog. Keep Reading!

Featured image Copyright of ‘The Human Face of Big Data’ Rick Smolan 2012-13.