Readings
TESTAR test results extracted while executing MyThaiStar as web system under test
Authors: Fernando Pastor Ricos and Tanja E. Vos from Universitat Politècnica de València
TESTAR test results datasets extracted with TESTAR tool using MyThaiStar web application as System Under Test (SUT). These datasets have been generated to be used as an example to be automatically generated and introduced locally in DECODER PKM, from H2020 DECODER Project.
TESTAR tool is an open source tool for automated testing through graphical user interface (GUI) currently being developed by the Universitat Politecnica de Valencia and the Open University of the Netherlands.
MyThaiStar is the reference application that Capgemini uses internally to promote best programming practices and the correct use of last technologies. It’s is developed with Devon Framework, the standard tool for development at the company. More...
An MLOps approach to bring models to production
Machine Learning Open Studio and Model as a Service (MaaS) from Activeeon helps data scientists and IT operations work together in an MLOps approach allowing to bring ML models to production. Machine Learning Open Studio includes automatic data drift detection mechanisms and allows traceability and audit over model performance to retrain it when necessary.
Only a small percentage of ML projects make it to production because of deployment complexity, lack of governance tools and many other reasons. Once in production, ML models often fail to adapt to the changes in the environment and its dynamic data which results in performance degradation.
To maintain the prediction accuracy of ML models in production, an active monitoring of model performance is mandatory. This allows to know when to retrain it using the most recent data and the newest implementation techniques, then redeploy in production. More...
Algorithm and Data Structure Visualization
Visualizations can help us understand how data structures and algorithms work.
The visualgo.net website provides great visualization and animations on advanced algorithms. Most of them are discussed in 'Competitive Programming', co-authored by two brothers Dr Steven Halim and Dr Felix Halim. Today, some of these advanced algorithms visualization/animation can only be found in VisuAlgo.
An online quiz system has been added that allows students to test their knowledge of basic data structures and algorithms. It generates questions and check the student answers automatically.
Covid-19 infection in Italy: when AI provides vital insights
Thanks to mathematical models and predictions, Gianluca Malato - a Data Scientist, fiction author and software developer - compared logistic and exponential models applied to Covid-19 virus infection in Italy. Both models help to better understand the evolution of the infection. The data preparation and python coding are detailed in an article posted in Towards Data Science on 8 March 2020. At that time, the main projections - now checked regularly by this Covid-19 Italian infection collaborative research - were:
- The expected number of infected people at infection end is 15968 +/- 4174.
- The infection peak is expected around 9 March 2020.
- The expected infection should end on 15 April 2020.
- Read the Covid-19 infection in Italy article in Towards Data Science
Clear Linux OS automates the creation of RPM packaging
Designed by Intel and open source contributors, the Clear Linux OS delivers a secure, hardware optimized OS. Its updates ensure that software dependencies remain mutually compatible.
The autospec tool is used to assist with the automated creation and maintenance of RPM packaging in Clear Linux OS. Where a standard RPM build process using rpmbuild requires a tarball and .spec file to start, autospec requires only a tarball and package name to start.
Recent reviews confirm the performance an stability improvements of Clear Linux OS. However, software that are packaged in other formats for other Linux distributions are not guaranteed to work on Clear Linux OS and may be impacted by Clear Linux OS updates.
The Twelve-Factor App, a Methodology for Building Web Apps
Suggested by the designers of the Heroku PaaS platform, the twelve-factor methodology can be applied to apps written in any programming language, and which use any combination of backing services (database, queue, memory cache, etc). It is aimed at building Software-as-a-Service apps that:
- Use declarative formats for setup automation, to minimize time and cost for new developers joining the project;
- Have a clean contract with the underlying operating system, offering maximum portability between execution environments;
- Are suitable for deployment on modern cloud platforms, obviating the need for servers and systems administration;
- Minimize divergence between development and production, enabling continuous deployment for maximum agility;
- And can scale up without significant changes to tooling, architecture, or development practices.
More about the Twelve-Factor App
A New Model-Based Approach for API Testing
Keeping Pace with Agile Development, Visualizing Complex Dependencies, and Orchestrating for Completeness of Testing are three good reasons to select a Model-Based approach for API testing, according to Collin Chau, a DevOps test expert.
"With the proliferation and complexity in microservices development that the Internet of Things brings, development teams are struggling to embrace API testing for more effective QA testing in-sprint. Learn how a model-based testing approach makes the difference in your API tests."
Read Collin Chau full article in Continuous Testing
NLP Search Paves the Way for Augmented Data Discovery
Combining natural language understanding and natural language generation will result in dynamic, bi-directional human-machine communication that will take several forms: text, voice and images. In text and voice scenarios, the BI or analytics solution can converse with the user to render the desired result - regardless of data-related and query-related search complexity.
Data visualizations also will become more interactive, if not immersive, along the lines of Busby from Oblong Industries. This product focuses on immersive interfaces, not specifically BI or analytics. However, its concepts could have a ripple effect on how people interact with data and thus, augmented data discovery.
"I think the future of BI is no BI. Don't ask me to search and look for things anymore. Give me that piece of information when I need it and if I need it. Come to me when there's something I need to know", foresees Erick Brethenoux, senior director analyst at Gartner.
For more information, read Lisa Morgan TechTarget article entitled NLP makes augmented data discovery a reality in analytics
Is BERT a Game Changer in NLP?
BERT (Bidirectional Encoder Representations from Transformers) is an open-sourced NLP pre-training model developed by researchers at Google in 2018. It has inspired multiple NLP architectures, training approaches and language models, including Google’s TransformerXL, OpenAI’s GPT-2, ERNIE2.0, XLNet, and RoBERTa.
For instance, BERT is now used by Google Search to provide more relevant results. And it can also be used in smarter chatbots with conversational AI applications, expects Bharat S Raj.
The information in this document is provided “as is”, and no guarantee or warranty is given that the information is fit for any particular purpose. The content of this document reflects only the author`s view – the European Commission is not responsible for any use that may be made of the information it contains. The users use the information at their sole risk and liability.
