Sentiment Analysis in Turkish Text with Machine Learning Algorithms

Merve Rumelli, Deniz Akkus, Ozge Kart, Zerrin Isik

Research output: Chapter in Book / Report / Conference proceedingConference contributionAcademicpeer-review

Abstract

With the developing technology, the number of comments made on the internet is increasing day by day. It has become almost impossible to make a manual sentiment analysis on these comments. Therefore, new algorithms should be developed to automatically perform sentiment analysis on these texts. In this study, a sentiment analysis model has been developed for Turkish texts. While developing this model, lexicon-based methods and machine learning algorithms were used together. As a naïve method of sentiment analysis, the root of each word in a sentence takes a score from a dictionary and the final polarity score of the relevant sentence is calculated by using additive score-based models. Machine learning models are trained to perform accurate sentiment annotations by using features based on polarity scores of texts. The final supervised machine learning model can achieve sentiment annotations of new Turkish texts within a 73% success rate without any human intervention.
Original languageEnglish
Title of host publicationProceedings - 2019 Innovations in Intelligent Systems and Applications Conference, ASYU 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728128689
DOIs
Publication statusPublished - 1 Oct 2019
Externally publishedYes
Event2019 Innovations in Intelligent Systems and Applications Conference, ASYU 2019 - Izmir, Turkey
Duration: 31 Oct 20192 Nov 2019

Conference

Conference2019 Innovations in Intelligent Systems and Applications Conference, ASYU 2019
Country/TerritoryTurkey
CityIzmir
Period31/10/192/11/19

Fingerprint

Dive into the research topics of 'Sentiment Analysis in Turkish Text with Machine Learning Algorithms'. Together they form a unique fingerprint.

Cite this