The National Information Processing Institute (OPI) has a surprise for anyone who wishes to develop their IT services and research projects: some of the institute’s IT tools are now available free of charge. Click on the ‘Download’ tab at http://www.opi.org.pl/ and choose a tool to use to your advantage. OPI offers machine learning models, natural language processing tools, open-access research datasets, and the source code for NAVOICA, the most popular e-learning platform in Poland. Users can access the tools without having to register or meet any formal requirements.
OPI champions the development of the information society
OPI is a modern centre of information technology and artificial intelligence. The motto of the institute is ‘from information to innovation’. OPI assists the Polish government in implementing science and higher education policy, as well as supporting Poland’s efforts to build an innovative, knowledge-based economy. The institute’s flagship IT tools include the POL-on Integrated System of Information on Science and Higher Education, the RAD-on integrated information platform, the ELA Polish Graduate Tracking System, the JSA Uniform Antiplagiarism System, the OSF Funding Stream Support System, the PBN Polish Scholarly Bibliography, the SEDN System for Evaluation of Scientific Achievements, and the NAVOICA e-learning portal.
‘Our staff are focused experts who specialise in information technology and other areas. We have been implementing IT projects and supporting Polish science for over thirty years. During this time, we have developed a large number of products—some of which have just been uploaded to our website. We have decided to share the results of our work with the public to speed up the development of the information society in Poland. All of the tools are available free of charge. We believe that the rapid growth of the IT sector and the scientific community in Poland will work to our advantage,’ said Dr Jarosław Protasiewicz, Head of OPI.
Open-access tools for natural language processing and research datasets
The algorithms that have been made available by OPI include open-access natural language processing tools. They rely on REST API and integrate four language models. The OPI-developed API enables users to train and test their own programmes based on natural language processing (NLP) solutions. The tool is simple, compact, and ready to use. It saves users time that would otherwise be required to configure multiple language models. Users can use preset components to quickly and seamlessly create their own advanced solutions and applications. OPI supports the IT sector and Polish science by making open-access research datasets available to the public. The data pertains to various OPI research projects. It can be used by any researcher, regardless of their affiliation.
NAVOICA and machine learning models
OPI’s machine learning models are some of the best in Poland and their number increases every year. Similar algorithms are used to translate text online and converse with chatbots. Most of the models are designed specifically for the English language. Experts at OPI have also made Polish-language models available free of charge to the public. Neural language models are being developed at many information centres around the world. The IT sector is fully aware of their potential; however, the development of new neural models requires considerable computational power, specialised infrastructure, and large datasets. For these reasons, they cannot be trained by individual researchers or small organisations. Like other tools that are based on artificial intelligence, neural models are much more accurate if they are trained on large volumes of data. One of the IT tools that has been made available by OPI is the Polish RoBERTa large model, which has been trained on the largest text corpus in Poland.
OPI has also uploaded the source code for the NAVOICA education platform, which is rapidly gaining in popularity. The portal boasts 85,000 registered users and has issued over 57,000 course completion certificates. NAVOICA is a learning management system (LMS) platform used to develop scalable e-learning massive open online course (MOOC) websites that enable the creation and implementation of courses for any number of participants in an asynchronous model. NAVOICA is a modified version of the Open edX platform. Experts at OPI hope that making the source code available to the public will help create new professional educational platforms.