Research

There are 44 posts filed in Research (this is page 1 of 4).

Some leisure time hacking…

After some time of working in a new public spending effort with my colleagues from the National Technical University of Athens, a Linked Open Data portal of public spending has been released. I have contibuted in the part of unifying company names and linking product scheme classifications, neverthless they have performed a very good job promoting to the Linked Open Data initiative a lot of public contracts metadata. See the next video to have an idea of the work:

Publicspending.net from John Fidias on Vimeo.

Hope to continue this fruitful collaboration!

After Prague…May!

Hi! after one month and a half some things have happened…I am going to summarize a bit…writing the “Highlights”!

Reading (pi = persona interest)

Reviewing

Writing

Coding and Tools

  • I have started to design a taxonomy collecting main KPI from the CSMIC initiative and other existing works to create the Cloudindex. Some initial outcomes can be checked here (it is very similar to my work at the Webindex, however the computation process must be added and the URIs are not yet accesible):
  • …I plan to report this on-going work to the CSMIC initiative!
  • I was also testing the IMPALA tool of Cloudera for real time queries.
  • I would like also to emphasize the excellent work of the Media Lab Team at the University of Athens creating the “PublicSpending.net” portal. My contribution has been small but I hope to contribute more in the future.

Teaching

  • During my stay in Spain in the Easter holidays (first week of may) I was invited by Jose Emilio Labra to give some seminars in the Semantic Web Course in the context of the Master of the Web Engineering. I am very grateful for this opportunity and I hope I can repeat. I was presenting part of my PhD and my current research SEERC. I think students were more or less happy with the contents! (My last lecturer had been in January I was a little bit not-trained).
  • Finally I have also collaborated in reviewing and partially supervising a final degree project on sentiment analysis. It has been a great experience!

Other things

In Spain I could meet my family and friends, visit places such as Tazones, go out and, mainly, drink and escanciar “Sidra Asturiana” and eat in places such as Tierra Astur or Los Lagos…and more things that you should ask me! 🙂

Map/Reduce intro

I have been playing with different Map/Reduce tools and examples to acquire the whole of view of this programming to solve different problems. Basically there are several patterns that can be identified when you need to tackle the implementation of some algorithm. I would like to leave some references that have served as inspiration:

In the same way I have prepared some slides compiling some basic examples that I have adapted from the original sources to deliver my own tutorial. I firmly believe that the best way of learning someting is try to explain to others so I hope to extend these slides in the future covering most of the patterns with different examples. At the moment I am going to
leave here the presentation:

Now it is time to play with Impala-Cloudera…I will keep you inform of the progress!

Week #12 and #13. The MSTC in Prague

Last week I was preparing my presentation for the RELATE-MSTC (Multi Skills Training Camp) in Prague. I was very exciting because it was going to be the first time I was to meet with my colleagues in the RELATE project. Now I can definitely say that the experience is being amazing! We have attended to  talks with different scopes: presentation skills, paralell programming, etc. In the same way we have attended to our own presentations. Each fellow was in charge of presenting their results, research questions and so on to finally keep a discussion with others.  This kind of collaboration meetings are very fruitful and I think I have improved my knowledge about some cloud computing issues. I have also presented my progress in quality of service and I hope to do my best in the next camp. Apart from that we are now attending to the ICPE 2013 conferencia, more specifically to the Hot topics workshop!

Let me leave my presentation  (position) here:


I will continue posting following the previous structure…

Papers to read

Reading on-going:

MapReduce Design patterns

Last and this week I have been preparing an introduction about the Map/Reduce algorithm I have found a lot of new excellent references (see my previous post) and I have read some books that I did not know. As a result I have made a compilation (it is just a summary that I will continue updating) that can serve as roadmap about what MapReduce is and what you can do with this programming model. As soon as I review my presentation and the examples I will upload them,

Week #10 and #11 in Thessaloniki

Hi! I have been a little bit lazy about writing in the blog but it is now time of recovering good practices. I am going to summarize my tasks during the last two weeks,

Reading (pi = persona interest)

Writing, reviewing and researching

  • I am managing a Special Issue in the Journal of Computers in Industry, Elsevier.
  • I finished the review of a book for Manning Publications.
  • I have been included as Technical Development Editor in Manning Publications.
  • I have been included as PC member in the workshop proposal “Data Mining on Linked Data (DMoLD’13) workshop with Linked Data Mining Challenge”  thanks to my colleagues at the University of Economics in Prague.
  • I am reviewing a paper for the journal “Expert Systems with Applications” (IF: 2.203)
  • I am reviewing and finishing the paper with my colleague Alejandro Montes about his final Master Project.

Meetings

  • I have had a meeting with my SEERC colleagues to talk about next actions.
  • I have had a meeting with Michalis Vafoupolus to prepare the Linked Data Cup paper.
  • I have had two meetings with Lum about his Bachelor Degree Project. It is a kind of supervising to address the problem of sentiment analysis using Rapidminer, Lingpipe, Alchemy API and a custom solution.

Coding and Tools

  • I have made in my leisure time a tool for unifying company names called CORFU using Python, NLTK and the APIs of Google Places, Linkedin and Google Suggestions. It also includes other algorithms based on string similarity, etc.
  • I have developed a simple sentiment analyzer using Alchemy API.
  • I have adapted some examples of Map Reduce patterns

Other things

  • I continue my fight to learn Greek…I have to study a bit more!
This is all I can remember…perhaps I am missing something…!

CFP Special issue on New trends on E-Procurement applying Semantic Technologies

Regarding the special issue in which I am a guest editor, please find the details in the next link.

Overview

E-Procurement refers to the use of electronic communications and transaction processing by government institutions and other public sector organizations when buying supplies and services or tendering public works. However, there is much more at stake than the mere changeover from paper based systems to ones using electronic communications for public procurement procedures. It should have the potential to yield important improvements in the efficiency of individual purchases, the overall administration of public procurement and the functioning of the markets for government contracts. The technology in this area may make it possible to automate the processes involved in the E-Procurement context besides features for supplier management and complex auctions should be included in by means of applying new technologies and methods in order to fulfill the requirements of this new realm of electronic businesses.

The new generation of E-Procurement is now on-demand or a software-as-a-service. There are seven main types of E-Procurement: 1) Web-based ERP (Enterprise Resource Planning)  creating and approving purchasing requisitions, placing purchase orders and receiving goods and services by using a software system based on Internet technology; 2)e-MRO (Maintenance, Repair and Overhaul), the same as web-based ERP except that the goods and services ordered are non-product related MRO supplies; 3) e-sourcing, identifying new suppliers for a specific category of purchasing requirements using Internet technology; 4)e-tendering, sending requests for information and prices to suppliers and receiving the responses of suppliers using Internet technology; 5) e-reverse auctioning, using Internet technology to buy goods and services from a number of known or unknown suppliers; 6)e-informing, gathering and distributing purchasing information both from and to internal and external parties using Internet technology; 7) e-market sites, expands on Web-based ERP to open up value chains. Buying communities can access preferred suppliers’ products and services, add to shopping carts, create requisition, seek approval, receipt purchase orders and process electronic invoices with integration to suppliers’ supply chains and buyers’ financial systems.

The features and requirements of an E-Procurement system match the advantages and features of software component and knowledge-based architectures. In this sense, Information technology support for the next generation of E-Procurement systems is critical since this information technology needs to be: (1) Reusable, and (2) Rapidly changeable. Furthermore, it can be seen that the underlying software architecture of a supply chain solution also needs to be: (1) Agile, (2) Flexible, (3) Deployable over a multi-enterprise scope, (4) Multi-function support when used as a suite, (5) Handle complexity, (6) Enable collaboration, (7) Enable coordination and (8) Greener to provide new green supply chain environments. Under this context, Semantic Technologies have emerged as an option to develop digital assistants and agents with the ability to search products and services on the Web that correspond best to the specific needs of a certain user. Semantic Web enables machines to interpret data published in a machine-interpretable form on the Web. The emerging Semantic Technologies have the potential to deeply influence the further development of the Internet Economy and support the different stages of e-Procurement processes such as eNotification, eAccess, eSubmission, eEvaluation/eAwarding, eOrdering, eInvoicing and ePayment.

In the Semantic context, Linked Data is one of the main technological cornerstones of Semantic Technologies for the improvement of E-Procurement. Linked Data describes a method of publishing structured data so that it can be interlinked and become more useful. Currently research works are arising from a semantic perspective in order to solve and improve the interoperability, privacy, trust,  quality, provenance, integration, transparency, lifecycle, etc. issues of this context. That is due the emerging Semantic technologies are open new development and business models in E-Procurement reusing the available information to deploy new innovative services of matchmaking, recommendation or tracking activities and trying to automate the processes involved in the E-Procurement sector. In that sense, existing solutions are taking advantage of these semantic approaches to deliver innovative and cost-effective information services. They are designed to add value to public sector information generated within the framework of public contracts published in EU Member States as well as the information generated by EU Institutions own activity.

The aim of this special issue is to collect innovative and high-quality research and industrial contributions regarding E-Procurement processes that can fulfill the needs of this new realm. This special issue aims at exploring the recent advances in the application of Semantic Technologies in the E-Procurement sector soliciting original scientific contributions in the form of theoretical, experimental and real research and case studies.

Submission Guidelines

Papers submitted must have not been published previously or under consideration for publication, though they may represent significant extensions of prior work. All submissions will be peer-reviewed following a double-blind review process. The objective is to apply very high standards of acceptance while ensuring fair, timely and efficient review cycles. The acceptance process will focus on those papers that address innovative contributions for adopting and implementing new techniques on E-Procurement applying semantic Web-based technologies.

Important dates and Timeline

  • 1st of May 2013,  full papers due (depending on the EES)
  • 15th of May 2013,  Chief editors desk review results
  • 1st of July, 2013, ending the first round of reviews
  • 15th of September 2013, revised papers due
  • 1st of November 2013, ending the second round of reviews
  • 1st of December 2013, final papers due

Week #9 in Thessaloniki

Just a few comments for this week (to be completed)….

Reading (pi = persona interest)

Writing, reviewing and researching

  • I am finishing the book chapter about publishing statistical data in RDF
  • I am managing a Special Issue in the Journal of Computers in Industry, Elsevier
  • I just realized that Labra added me in the Acknowledgements part of his work about “Multilingual Open Data Patterns” I am very proud of that! (to be honest I just collaborated in the first presentation with some links and specially through some comments with regards to SKOS-XL). I also suggest to read the paper in which each of the patterns is explained and discussed with excellent examples.

Meetings

  • I have had a meeting with my SEERC colleagues to present my prototype and plan next actions in QoS, etc.
  • I have had a meeting with Michalis Vafoupolus to prepare the Linked Data Cup paper.

Coding and Tools

  • I have implemented a real-time based architecture using the Lambda approach and following some hints from Pere Ferrera. It is not the same algorithm and I am just take the approach to tackle the problem not source code. Next steps include to use RDF as views for batch and real-time layers using SPARQL federated queries (for instance Fedex). The example just takes a Twitter stream using Tweet4J API and counts words presenting the results in a HTML page. Documentation is available here and also the the source code (under development).
  • I have linked to the CPV the public procurement notices from UK, USA and AUS.

Other things

Week #8 in Thessaloniki

Just a few comments for this week (to be completed)….

Reading (pi = persona interest)

Writing, reviewing and researching

Meetings

  • I have had a meeting with my SEERC colleagues to present my prototype and plan next actions in QoS, etc.
  • I have attended a call with the Oviedo Council (planned)

Coding and Tools

  • I have re-implemented my real time word counter of Twitter status using RabbitMQ, Node.js, Storm and Redis (see Demo wiki page).
  • I am going to test the MOA framework (based on Weka) for mining data streams. In this sense I have also found an implementation of MOA in Storm.

Other things

I have Internet connection at home!

Week #7 in Thessaloniki

Just a few comments for this week (to be completed)….

Reading (pi = persona interest)

Writing, reviewing and researching

  • I am reviewing a paper for a Special Isuee of a JCR journal
  • I am finishing the book chapter about publishing statistical data in RDF
  • I have also made the first review of WESOMENDER (we have to work hard to get a good contribution but the expectations are high)
  • I have been invited to be part of the PC of the Special Session “Engineering Tool Integration for Industrial Automation System Development (ETAS 2013)” in conjuction with IECON2013
  • I have joined in the research group “Comercio Electronico en Colombia – GICOECOL” thanks to Luz Andrea RODRIGUEZ ROJAS with whom I will collaborate to empower the use of Open Data in e-Health.

Meetings

Coding and Tools

I have implemented a real time word counter of Twitter status using different tecniques:

  • The classical Observer design pattern
  • The Storm framework, I have reused some examples to implement my own spouts and bolts
  • The Trident framework on the top of Storm, I have also reused some examples of the storm-starter project customizing the code to get a better understanding

Other things

This week I have started the 3-month Greek course and I am very happy because I can now understand some words and read a little bit 🙂 Besides my classmates are from a lot of countries: Bulgaria, Germany, Bosnia, France, Serbia, New Zealand, Italy, Moldova and Russia. It is a GREAT experience.

 

Week #6 in Thessaloniki

Just a few comments for this week….

Reading

Writing and reviewing

  • I have written a paper entitled “Semantic-based QoS management in Cloud Systems: Current Status and Future Challenges “. It is a survey of 25 pages in which I review most of the approaches for QoS in Cloud Computing as well as the main techniques for stream reasoning in the Semantic Web area. It must be the kick-off of the architecture for semantic-based QoS management in Cloud Systems. and I also have the acronym “SEQoS” (remember it because you will hear about it!)
  • I have also managed tasks regarding the special issue in COMIND.

Coding and Tools

I have not made any relevant progress in developing tasks.

Week #5 in Thessaloniki

The last week I have been focused on two main tasks: my presentation at the City College and the submission of a paper. Following the same structure of last weeks I leave some links to the activities I am carrying out:

Reading

Writing and reviewing

  • I have continued with the structure and firts contents of two papers and one special issue proposal.
  • I have managed all the abstracts for the COMIND Special Issue.
  • I have submitted a paper to “Computers and Human Behavior“
  • I have made the presentation in the following bullet to the Deparment of Computer Science at City College
  • I have reviewed my previous presentation about MOLDEAS and the new one is supposed to be more didactic such as an “Intro” to Linked Open Data

Coding and Tools

I have not made any relevant progress in developing tasks.