Two weeks ago I attended the 3rd International Conference on Cloud Computing and Services Science, CLOSER 2013 to present a poster about cloud computing services management. I’ve detailed the topic of the poster on my previous blog post available here.
The conference took place in Eurogress, Aachen. Actually, I had the chance to follow the other conferences held in conjunction with Closer:
WEBIST 2013
This conference is close to my research activities in that it focuses on the use of web-technologies in business applications. I had fruitful discussions with researchers working on issues such as integration of social networks [1], collaborative BI [2] , and the similarities between XML and graph management [3].
SMARTGREENS 2013
Green IT is becoming a trendy topic focused on lowering energy consumption, and by the way cost reduction, of computing infrastructures as it starts to be more demanding (data centers) and ubiquitous (mobile devices). For example, ICT consumes 10.5% of energy in Germany (Deutscher Bundestag, 2010), and produces 2% of worldwide CO2 emission and rising – equivalent to aviation industry (Gartner, 2007).
CSEDU 2013
e-learning is a hot topic gaining more interest both from academia and industry. The field lies on the intersection of computer science and pedagogy. Teaching online brings its new challenges. With thousands of students from all over the world, and thousands of access to the platform either for following the courses, filling the home work or taking the exams. Among the interesting research directions tackled by this community is the scale, course materials, students assessment, fraud detection, diploma equivalence etc. The challenges are big, however it is attracting more attention. Coursera is a great example of a successful industrial application of such research field. It was elected as the best new startup of 2013 [4].
Big Data
Big Data is unsurprisingly a hot topic of the conference. And a cloud infrastructure is definitely the right infrastructure to handle the 3V aspects, even more!
However, we should keep in mind that better algorithms are as important as better hardware when it comes to big data management.
An attempt to define BigData ecosystem was presented at a keynote on Closer, the statement is as follows:
“Use Clouds running Data Analytics expressed as Services processing Big Data to solve problems in X-Informatics.”
In 2013, world hard drive production is about 20 exabytes/day, while data is generated at a rate of 5 exabytes/10min. A solution to this point was discussed on a keynote given on Webist, by Software AG. The subject is the integration of complex event processing (CEP) with current data processing workflows.
Using CEP, data processing is done using static queries on changing data, the opposite paradigm of SQL-based system. In a nutshell, big data is still growing and not all of it is worth persisting, hence it becomes interesting to manage it in a process and forgot fashion. Operations such as pattern matching are applied for predefined time-windows on the data stream. Once the time-window is over, the data is discarded, and only the analysis results are persisted.
Closer 2013
Up to the clouds, SLA management, privacy&security, standards and monitoring are as usual the key research directions on the cloud community. Our contribution was based on cloud computing open standards. Even though I agree that standards suffers from the weakness that to be adopted they should be pushed by big cloud players such as Amazon, Google and Microsoft. The process reminds me the SQL language, to this day none of the vendor has completely compliant implementation to the defined standard [5].
Back to green IT, even for the cloud community, a real interest on the energy consumption is shown on many of the talks. The best paper award was attributed to the paper about: Characterising the Power Consumption of Hadoop Clouds [6].
The conference was also a great moment to share our research perspectives with researchers from multiple labs, and potentially start new collaborations with teams working on the data processing and management field.
Job Opportunities!
The last point I would like to mention is jobs opportunities brought by clouds and big data. According to Microsoft, 14 million cloud jobs are expected on 2015. McKinsey predicts 190,000 nerds and 1.5 million extra manager needed in Data science by 2018 in USA only [7]!
For more details about any of the above topics, don’t hesitate to comment or email me!
Amine Ghrab
Twitter: @AmineGhrab
References
[1] Xuan-Truong Vu, Pierre Morizet-Mahoudeaux and Marie-Hélène Abel. User-centered Social Network Profiles Integration. Webist2013
[2] Jens Kaufmann and Peter Chamoni. Empowering Collaborative Business Intelligence by the use of Online Social Networks. Webist2013
[3] Stefan Böttcher, Marc Brandenburg and Rita Hartel. XML DAG – Index – A Compressed Index for XML Keyword Search. Webist2013
[4] http://techcrunch.com/2013/01/31/coursera-wins-best-new-startup-of-2012-get-schooled/
[5] http://en.wikipedia.org/wiki/SQL#Criticism
[6] Javier Conejero, Omer Rana, Peter Burnap, Jeffrey Morgan, Carmen Carrion and Blanca Caminero. Characterising the Power Consumption of Hadoop Clouds – A Social Media Analysis Case Study. Closer2013
[7] http://www.mckinsey.com/insights/business_technology/big_data_the_next_frontier_for_innovation