The next online scientific seminar of the 1st department was held at the Institute of Information Technology of ANAS.
Senior researcher of the institute Marziya Ismayilova made a report on "Sentiment analysis: problems and existing approaches", the history and stages of development of sentiment analysis, main issues, levels, classification methods, areas of application, problems, software products, online tools, etc. gave detailed information about.
The speaker noted that sentiment analysis is a matter of discovering, extracting, and classifying the emotional content of text from text documents using natural language processing (NLP), statistics, or machine learning methods.
According to Ismayilova, sentiment analysis is also known as opinion mining or emotion artifical intelligence. It is based on the use of natural language processing, text analysis, computer linguistics and biometrics for the systematic recognition, extraction, measurement and study of emotional states and subjective information. Sentiment analysis is the process of automatically determining the author's emotional response to an object by content analysis methods. It is a matter of automatic detection and classification of ideas expressed in texts written in natural language.
"The history of sentiment analysis is based on the analysis of public opinion in the early twentieth century and the analysis of text subjectivity in the field of computer linguistics in the 1990s," she said.
She presented a graph showing the number of articles on sentiment analysis published in the Web of Science and Google Scholar databases between 2002 and 2019. Then, the speaker presented a conceptual model of sentiment analysis and provided information on the main terms of sentiment analysis and sources of information.
She also spoke about the types of sentiment analysis based on characteristics, its words and their frequency of use, term position, parts of speech, feedback words and phrases, negative words, syntax dependence, based on vocabulary and rules for sentiment analysis, as well as machine stated that classification methods were used as existing approaches to training methods. The speaker presented a model of sentiment analysis and noted that sentiment analysis is applied in such fields as sociology, medicine and psychology, marketing, political science.
Ismayilova spoke about the Twitter Sentiment web service used for sentiment analysis, software products such as SA with Python NLTK TC and I-Teco, and the problems of sentiment analysis.
She provided detailed information on online tools, books on sentiment analysis, conferences, symposia and workshops, as well as research scientists working in this field.
There was an exchange of views on the report, questions were answered.