The Center for Strategic Analytics and Big Data held an autumn series of webinars on methodological innovations of iFORA big data intellectual analysis system
During this series of events, the Centre’s employees shared their experience in applying new methodological solutions in big data analysis. Additionally, the developers of the iFORA big data intelligent analysis system demonstrated its effectiveness using real cases. The webinars took place on November 13, 15 and 18, 2024.
Overall, three webinars were held.
1. How to Improve the Quality of Big Text Data Analysis? Using the Example of a Comparison of Global and Chinese Trends in Digitalisation of Education
On November 13, expert Maria Antasheva and leading programmers Yulia Isaeva and Evgenia Zakovorotnaya demonstrated new trends in term vectorization based on the principle of analyzing Chinese trends in education. The approach showed high quality metrics and led to a significant increase in the results of semantic analysis. This was demonstrated by the example of identifying global trends in the digitalisation of education: the relevance of the results to the search query has increased, and the comparison of global and individual countries’ agendas (China was chosen as a case study) has shown the existing differences. For example, the influence of digitalization of education on the digital economy of the country has been manifested in China’s agenda, and a special emphasis is also noticeable when it comes to training talents in the field of information technology.
2. What Is Named Entity Recognition and Why Analysing Chinese Names Is So Tricky?
On November 15, during the second session, expert Maria Antasheva and leading programmer Yulia Isaeva spoke about the Named Entity Recognition (NER) algorithm used to identify names, organisations’ titles and other proper names in Chinese texts. Named Entity Recognition on a large-scale dataset can be challenging in general, but when it comes to working with unstructured texts in the Chinese language it becomes especially tricky.
The ISSEK researchers tested the algorithm on Chinese texts on science and education. The iFORA system could reveal not only the names of scientific and educational organisations but also areas of learning and development in educational programmes as well as the names of research centres created jointly with foreign partners. This and other valuable insights are now available thanks to the new methodology.
3. iFORA 2.0: How to Improve the Security and Efficiency of the Big Data Analysis System Infrastructure?
Finally, on November 18, programmer Nikita Savvin and expert Maria Antasheva shared with the webinar guests insights about the work that is being done to improve the security and efficiency of iFORA. The efficiency of the entire big data analytics system depends on how data storage, processing algorithms and infrastructure load monitoring are organised. This year, iFORA system developers have improved approaches to data storage to increase access security, and load management for ever-increasing data volumes, and updated a number of solutions that optimise the system’s operation and monitoring function. In this webinar, they will discuss the major challenges they faced in building and implementing the new infrastructure.
Overall, the webinar series brought together more than 100 listeners from Russia, including guests from Moscow, St. Petersburg, Perm and Nizhny Novgorod, as well as from abroad (Kazakhstan, Bangladesh). A series of open lectures on the use of big data analysis in ISSEK research activities is being held as part of the World-Class Research Centres (WCRCs) project "Center for Interdisciplinary Research of Human Potential".