The first semantic map was generated at the end of 2018 when we decided to process Wikipedia articles, please find more details in Semantic map for Wikipedia.
The results could hardly leave anyone indifferent: we got a vivid picture of the entire human knowledge. And this picture can be explored infinitely. Each area of the map has its own meaning. You can understand the knowledge contained in the map by exploring the neighboring regions and see how the keywords gradually change, passing into other areas. You can meet rather unexpected topics and trace how they are connected.
We processed medical articles, papers on economics, and legal documents after that. Semantic maps made it possible to discover quickly which documents are most important, revealing relations between topics, sometimes intuitively clear and sometimes unexpected.
The same technology was used to analyze the dataset of COVID-19 related papers. The COVID-19 Open Research Dataset (CORD-19), was recently released by a group of research institutions; it contains more than 59 000 full-text research papers, over 1 Gb of text in total. We used the same technology to analyze this dataset.
The semantic map consists of several interfaces, which are further explained in the following sections: