Akrivi Krouska

The effect of preprocessing techniques on Twitter sentiment analysis

As Twitter offers a fertile ground for expressing different thoughts and opinions, it can be seen as a valuable tool for sentiment analysis. Furthermore, properly identified reviews present a baseline of information as an input to different systems, such as e-learning systems, decision support systems etc. However, the data preprocessing is a crucial step in sentiment analysis, since selecting the appropriate preprocessing methods, the correctly classified instances can be increased. In view of the above, this research paper explains the necessary information to get preprocess the reviews in order to find sentiment and make analysis whether it is positive or negative. Extended comparison of sentiment polarity classification methods for Twitter text and the role of text preprocessing in sentiment analysis are discussed in depth. In the set of tests, possible combinations of methods and report on their efficiency were included, conducting experiments using manually annotated Twitter datasets. Finally, it is proved that feature selection and representation can affect the classification performance positively.

Evaluation of ensemble-based sentiment classifiers for Twitter data

Social media are widely used worldwide and offer the possibility to users to post real time messages respecting their opinions on different topics, discuss everyday issues, complain and express positive, neutral or negative sentiments for anything that concerns them. As such, sentiment analysis has become a burning issue in the scientific literature. However, some researchers argue that Twitter sentiment classification performance may be elusive. To overcome this issue, in this paper, we evaluate the most common ensemble methods that can be used for effective sentiment analysis and the tested datasets used in this research proceed from Twitter. Experiment results reveal that the use of ensembles of multiple base classifiers can improve the accuracy of Twitter sentiment analysis. The discussion that is presented can clearly prove that such methods can surprisingly surpass the traditional algorithms in performance and can be seen as a beneficial tool in the field of sentiment analysis that can further enhance several other domains such as e-learning and web advertising.

Trends on Sentiment Analysis over Social Networks: Pre-processing Ramifications, Stand-Alone Classifiers and Ensemble Averaging

Technology advancements gave birth to social networks during the last decade. Many people tend to increasingly use them in order to share their personal opinion on current topics of their everyday life as well as express their emotions about situations in which they are interested. Hence, the emotions that are expressed in social networks can be positive, negative or neutral. To this direction, the analysis of people’s sentiments has drawn the attention of many scientists worldwide and offers a fertile ground for increasing research. Furthermore, a social network that is adopted by an ever growing percentage of people is Twitter. Twitter is an online news and social networking service where users post and interact with messages; such messages conceal people’s feelings and sentiments. Therefore, Twitter can be seen as a source of information and holds a vast amount of data that can be exploited for sentiment analysis research. In view of above, the purpose of this paper is to provide a guideline for the decision of optimal pre-processing techniques and classifiers for sentiment analysis over Twitter. In this context, three well-known Twitter datasets (OMD, HCR and STS-Gold) were used and a set of experiments was conducted. In particular, firstly, an extended comparison of sentiment polarity classification methods for Twitter text and the role of text preprocessing in sentiment analysis are discussed in depth. Secondly, four well-known learning-based classifiers (Naive Bayes, Support Vector Machine, k-Nearest Neighbors and C4.5) have been evaluated based on confusion matrices. Thirdly, the most common ensemble methods (Bagging, Boosting, Stacking and Voting) are examined and compared to base classifiers’ results. Finally, a case study concerning the application of Twitter sentiment analysis in an e-learning context is presented. The main result of the utilization of the Twitter-based learning application is that the exploitation of students’ emotional states can be used to enhance adaptivity in the learning content as well as deliver recommendations about activities and provide personalized assistance. Concerning data pre-processing, the experimental results demonstrate that feature selection and representation can affect the classification performance positively. Regarding the selection of the proper classifier, the superiority of Naive Bayes and Support Vector Machine, regardless of datasets, is proved, while the use of ensembles of multiple base classifiers can improve the accuracy of Twitter sentiment analysis.

Computerized Adaptive Assessment Using Accumulative Learning Activities Based on Revised Bloom’s Taxonomy

The need of developing more efficient educational systems leads to the incorporation of personalized operations. Digital learning focuses mainly on the adaptive content and navigation. However, the provision of a valid assessment tool is essential to an integrated e-learning system, as it indicates the accomplishment of learning goals. One such testing model should be designed based on the new demands to knowledge, skills and attitudes that students have to acquire, and considering the students’ needs. To this direction, this paper introduces an adaptive assessment system where the test items are designed based on Revised Bloom’s Taxonomy and the assessment content is adapted to students. One major advantage of the proposed system is that it provides a better detection of student’s learning gaps, useful for further system adaptivity.