• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

News

Method of semantic analysis and semantic networks

On November 29, Olga Silyutina and Alina Bakhitova gave presentations on methods of semantic analysis and semantic networks as part of the regular seminar of the Research Group "Machine Learning and Social Computing". 

The participant of our Research Group Alina Bakhitova presented the basics of the semantic analysis. Her report was conducted in the form of the tutorial, so that interested participants could use their laptops and follow the code. Among the participants of the seminar were also the 2nd and 3rd-year students of BA Sociology and Social Informatics. Anyway, text analysis and the allocation of sentiments were interesting for everyone. Many have thought about its application in their own research, for example in the project on eSports and highlighting opinions.

Next, Olga presented semantic networks in AutoMap. Since Alina's tutorial was based on the R programming language, many people liked the possibility of semantic analysis without "coding". To begin with, Olga compared the program with the familiar to students VosViewer, what helped to understand AutoMap much faster.

The thematic analysis and the allocation of sentiments are already being used by some participants of our Research Group in their projects.  For example, Viktor Karepin, Nikita Kaspruk and Olga Silyutina analyzed the reviews on St. Petersburg hotels, taking into account the sentiment, which allowed finding quite interesting results. This topic is also useful for analyzing reputation systems, social networks, and can also be an important part of creating rating models.