Feature based sentiment analysis on informal Turkish texts

2016
Kama, Batuhan
Show full item record
122
views
0
downloads
Sentiment analysis (SA); in other words, opining mining, is the automatic extraction process of a author' s feeling on a specific topic. These feelings can be positive, negative or neutral. However, most of the time authors do not carry the same opinion for all parts of a topic. While they have a positive attitude on some parts of a topic, they can criticise other parts. Therefore, feature or aspect based sentiment analysis (FBSA) a specialized version of sentiment analysis is used to analyse people' s attitude on each specific feature of a topic. Nowadays, there are a lot of forums, blogs and review sites where customers and professional reviewers add their comments on different products. Since thousands of new entries are written to these sites each day, it is almost impossible to read all comments on a product and its features and learn powerful and weak sides of this product; and hence, the need for classifying these informal online data automatically has increased and FBSA can be used to fulfill this need. Up to date, most of the studies on both sentiment analysis and feature based sentiment analysis was on English. There are only several works on these topic on Turkish. In this thesis, a dataset created by collecting comments and reviews from forums is processed with existing feature based sentiment analysis methods for English. Moreover, new methods for Turkish feature based sentiment analysis is proposed. 
Sentimentalism., Turkish language., Text editors (Computer programs)., Natural language processing (Computer science).
http://etd.lib.metu.edu.tr/upload/12620316/index.pdf 
https://hdl.handle.net/11511/25968
Graduate School of Natural and Applied Sciences, Thesis

Suggestions

Investigating the performance of segmentation methods with deep learning models for sentiment analysis on turkish informal texts
Kurt, Fatih; Karagöz, Pınar; Department of Information Systems (2018)
This work investigates segmentation approaches for informal short texts in morphologically rich languages in order to e ectively classify the sentiment. The two building blocks of the proposed work in this thesis are segmentation and deep neural network model building. Segmentation focuses on preprocessing of text with di erent methodologies. These methodologies are grouped under four distinct approaches; namely, morphological, sub-word, tokenization, and hybrid approaches. There is mostly multiple numbers ...
Extracting Potentially High Profit Product Feature Groups by Using High Utility Pattern Mining and Aspect Based Sentiment Analysis
Demir, Seyfullah; Alkan, Öznur; Çekinel, Recep Fırat; Karagöz, Pınar (Springer, Cham, 2019-01-01)
As a subproblem of sentiment analysis topic, aspect based sentiment analysis aims to extract distinct opinions for different aspects of a case in a given text. When the case is product review, it is possible to understand reviewer’s opinion on features of the product, rather than the product in general. Then, a product feature can be associated with a sentiment score denoting user satisfaction value by that feature. Modeling features mentioned in a review as items in a transaction may provide better insight...
Automatic sense prediction of implicit discourse relations in Turkish
Kurfalı, Murathan; Zeyrek Bozşahin, Deniz; Department of Cognitive Sciences (2016)
In discourse parsing, the sense prediction of the Implicit discourse relations poses the most significant challenge. The thesis aims to develop a supervised system to predict the sense of implicit discourse relations in Turkish Discourse Bank (TDB). In order to accomplish that goal, the discourse level annotations obtained from TDB are used. TDB follows the PDTB-2’s sense hierarchy and for all experiments within the current study, only CLASS senses are considered. As the primary experiment, the classifiers ...
Timing and X-ray spectral features of Swift J1626.6-5156
Icdem, Burcin; İNAM, SITKI ÇAĞDAŞ; Baykal, Altan (Oxford University Press (OUP), 2011-08-01)
In this paper, we extend the timing analysis by Baykal et al. of Swift J1626.6-5156 using RXTE-PCA observations between MJD 53724 and MJD 55113 together with a Chandra-ACIS observation on MJD 54897 with a 20-ks exposure. We also present the X-ray spectral analysis of these RXTE and Chandra observations. We find that the spin-up rate of the source is correlated with the X-ray flux. Using this correlation, we estimate the distance and surface magnetic field of the source as similar to 15 kpc and similar to 9 ...
Sentiment analysis in Turkish
Eroğul, Umut; Alpaslan, Ferda Nur; Department of Computer Engineering (2009)
Sentiment analysis is the automatic classification of a text, trying to determine the attitude of the writer with respect to a specific topic. The attitude may be either their judgment or evaluation, their feelings or the intended emotional communication. The recent increase in the use of review sites and blogs, has made a great amount of subjective data available. Nowadays, it is nearly impossible to manually process all the relevant data available, and as a consequence, the importance given to the automat...
Citation Formats
B. Kama, “Feature based sentiment analysis on informal Turkish texts,” M.S. - Master of Science, Middle East Technical University, 2016.
METU IIS - Integrated Information Service open@metu.edu.tr