Sentiment Analysis in Latvian and Russian: A Survey

Abstract

Social networking sites such as Facebook, Twitter and VKontakte, online stores such as eBay, Amazon and Alibaba as well as many other websites allow users to share their thoughts with their peers. Often those thoughts contain not only factual information, but also users’ opinion and feelings. This subjective information may be extracted using sentiment analysis methods, which are currently a topic of active research. Most studies are carried out on the basis of texts written in English, while other languages are being less researched. The present survey focuses on research conducted on the sentiment analysis for the Latvian and Russian languages.

The Applications of Sentiment Analysis for Russian Language Texts: Current Challenges and Future Perspectives

“…Koltsova and Nagornyy examined which topics were defined as social problems by reviewing the comments of the audiences of the regional mass media in Russia [57]. They collected a dataset of 33,887 news items and 258,107 comments from local online media sources in Omsk (Gorod55 9 , BK55 10 , NGS Omsk 11 , and Omsk-Inform 12 ) during the period from September 2013 to September 2014. To extract topics from news texts, the authors applied the Gensim [117] implementation of the Latent Dirichlet Allocation algorithm [102] with a metric by Arun, Suresh, Madhavan and Murthy [118].…”

“…Only one survey [10] by Viksna and Jekabsons directly addressed the sentiment analysis of Russian content, and several others [11]- [14] mentioned sentiment analysis of Russian in the contexts of overall comparison with globally existing approaches. There were other studies dedicated to specific aspects of sentiment analysis of Russian, for instance, the evaluation of state-of-the-art approaches [15]- [18], comparison of neural network architectures for sentiment analysis [19], [20], and comparison of publicly available Russian sentiment lexicons [21].…”

What Can We Learn from Almost a Decade of Food Tweets

“…The corpus is available on Github 4 , in accordance with the content redistribution section of the Twitter Developer Agreement and Policy 5 . The public release includes tweet IDs along with data fields created within the scope of this project (starting with "location_lng" in Figure 1).…”

“…We compared a Python implementation of the Naive Bayes classifier from NLTK [14] against Pinnis [3] implementation of the Perceptron classifier. We also experimented with several combinations of training data sets -TE (our Twitter Eater dataset), MP [3], RV [5], PE [4], NI 11 . We found that the highest classification accuracy -61.23 % -is achieved by using all but NI data sets for training and only stemming all words.…”

“…Milajevs [1] collected and analysed 1.4 million tweets geo-located in Riga, Latvia from April 2017 to July 2018 and 60 thousand tweets [2] from November 2016 to March 2017. Pinnis [3] collected and analysed 3.8 million tweets of Latvian politicians, companies, media, and users who interacted from August 2016 to July 2018 There are also several data sets of general sentiment-annotated tweets [4], [5], [3] 3 amounting to 14,781 tweets in total.…”