That was it!
When I started this blog I was thinking of it as a way to interact with others on the topic of my PhD thesis. As many things in life, this turned it into something else (not sure what).
But, in order to keep the initial mission close… this post is as a marker to the date in which I publicly presented my research and I was accepted to bear the title “Doctor of Philosophy“. And I must say I am proud.
I’m proud I did not give up during all these many years and that I listened to my teacher, wife and friends and continued on the road. I’m proud I learned a painful lesson in “backup” which triggered a complete rewrite of the thesis just before it was done. I’m proud the research is my own work, although imperfect or reasoning to others (hey! very few ideas are completely new in IT).
Published works? Just a few. You will find them in the bibliography of the paper. Which should be published online any day now by the university. Link to it? Maybe.. I’m trying to keep this blog as free as possible. Free as in speech, so I find comfortable the fact that my name is not directly related to it. I might be acting stupid here, but I’ll keep it Web crawler anonymous for the time being.
For you humans out there, reading this stuff… It means you care about the subject. So, for the sake of it, here is the paper abstract:
Structural Synthesis of High Performance DatabasesThe increasing volume of unstructured data existent inside current organizations generates the need for its efficient management. Databases are usually the informational support used to manage such volumes. This paper will present studies and experiments for both the design and the implementation of data having considerable size – content management systems.While content management tasks can be easily implemented using the existing database systems, we believe that high performance implementations require specialized processing and architectures.We will present the characteristics of content management systems, focusing on its usage as a data management platform. One of the most important sections of the paper will define the performance metrics and the most important challenges generated from these performance expectations. Performance is defined not only as a quantitative measure of response time but also from other perspectives such as the system resilience in long time periods (tens of years) given the continuous technical evolution.A new architecture is proposed in the paper, following the analysis of the main identified challenges. An actual implementation of the architecture is also described and several important design decisions are detailed. The implementation quantitiative performance behaviour is then assesed in order to validate the architectural decisions and observer the scalability to large data volumes. The implementation is then benchmarked in comparison with an alternate implementation of the same architecture but using an off the shelf standard database management system.The observed performance tests show that the architecture allows high performance metrics to be achieved and that it compares very well to other common database management systems and therefore we consider that we succesfully designed a technical implementation of a new model which supplies greater performance than conventional implementations.As next steps we aim to implement a standard interface to the designed system, using the CMIS (Content Management Interoperability Standard). This will open the possibilities to test the proposed system with third party software applications and will give a more precise indication of how it compares with existing traditional products from a performance point of view
Puzzled by the title? Don’t be. Back in 2000 when this title was conceived we thought the term “content management” or “ECM” would not sound academic enough to the community. Inside it’s all about a CMS engine (in the broader sense, not limited to web content management).
So long now, I go get some sleep and followup here shortly.