Internet as a corpus - a new seminar in the ABBYY Open series

Next Tuesday, July 19, a regular seminar in the ABBYY Open series “Actual Problems of Computational Linguistics” will take place in the Moscow office of ABBYY. Sergey Sharov, an employee of the Department of Translation at the University of Leeds (UK), who previously worked at the Russian Research Institute of Artificial Intelligence and the Russian Language Institute, RAS, will speak at the seminar. His report “Web as Corpus, Approaches to the quantitative and qualitative analysis of the textual content of the Internet” is devoted to methods of collecting linguistic corpuses on the Internet, assessing the quality of these methods and examining approaches to automatic text classification.

The seminar will describe ways to quickly collect cases in the desired area, approaches to automatic text classification by subject areas and genres using methods such as Support Vector Machines (SVM), Topic Modeling, Multidimensional Scaling. In addition to the quantitative assessment of the quality of methods, it is also necessary to carry out a qualitative assessment of the conformity of the results of the classification of language intuition. The seminar will provide examples of the use of methods for creating and processing buildings for Russian, English, Chinese and German.

Detailed information about the event you can read here . The seminar is free, you must register and wait for confirmation of registration.
')
UPD: Video from the seminar can be found here.