Natural Language Processing

Social media data analysis:

Author profiling, Stance detection, Deceptive opinion detection, Irony detection and sentiment analysis, Mixed-script text analysis, Plagiarism and social copying detection.


  • Author profiling

    Given a text, what are the author’s traits? The focus is on inferring traits such as gender, age, native language, language variety, and personality on the basis of the stylistic analysis of the author’s texts. This is of interest for areas such as marketing, forensics or security.


  • Stance detection

    Given a particular target entity, is the author’s opinion in favour or against? The focus is on monitoring sentiment in political debates or controversial topics to understand the way polarized communities communicate.


  • Deceptive opinion detection

    Given a review, is the author’s opinion truthful or the text was deliberately written to sound authentic in order to deceive the reader? The focus is on detecting fake reviews written to change purchase decisions and opinions about products and services.


  • Irony detection and sentiment analysis

    Given an opinion, is the author’s ironic? The focus is on understanding the opinion’s polarity in case forms of figurative language such as irony or sarcasm are used.


  • Mixed-script text analysis

    Given a query written in a given script, is it possible to retrieve texts written in mixed scripts? The focus is on handling cross-script term matching.


  • Plagiarism and social copying detection

    Given a text, is it an original? The focus is to detect the source text plagiarism is committed from.