Albanian Text Classification: Bag of Words Model and Word Analogies

Abstract

Background: Text classification is a very important task in information retrieval. Its objective is to classify new text documents in a set of predefined classes, using different supervised algorithms. Objectives: We focus on the text classification for Albanian news articles using two approaches. Methods/Approach: In the first approach, the words in a collection are considered as independent components, allocating to each of them a conforming vector in the vector’s space. Here we utilized nine classifiers from the scikit-learn package, training the classifiers with part of news articles (80%) and testing the accuracy with the remaining part of these articles. In the second approach, the text classification treats words based on their semantic and syntactic word similarities, supposing a word is formed by n-grams of characters. In this case, we have used the fastText, a hierarchical classifier, that considers local word order, as well as sub-word information. We have measured the accuracy for each classifier separately. We have also analyzed the training and testing time. Results: Our results show that the bag of words model does better than fastText when testing the classification process for not a large dataset of text. FastText shows better performance when classifying multi-label text. Conclusions: News articles can serve to create a benchmark for testing classification algorithms of Albanian texts. The best results are achieved with a bag of words model, with an accuracy of 94%.

Natural language processing for Albanian: a state-of-the-art survey

“…Later on, Kadriu et al [24] presented an extended study concerning text classification for Albanian news articles, which was based on a bag of words model and word analogies. In this study, the authors focused on text classification using two approaches: i) the text classification treats words as independent components and ii) the text classification treats words based on their semantic and syntactic word similarities.…”

Sentiment Analysis of Customer Feedback in Online Food Ordering Services

“…(1) Based on machine learning (Kadriu et al, 2019;Khairnar et al, 2013;Le et al, 2017) and (2) based on lexicon (Liu, 2012(Liu, , 2017Vu et al, 2017;Li et al, 2019). In addition, to increase the efficiency of the opinion mining method, the research has used a hybrid method combining machine learning and lexicon (Mudambi et al, 2010;Maks et al, 2012;Sun et al, 2017;Yang et al, 2017).…”

“…Due to some characteristics of the language on social networks, such as a limited number of characters or emotions depending heavily on what users are reading and listening to, the emotional classification of users in social networks is a challenging issue. Machine learning has been applied and has achieved some success in sentiment analysis (Khairnar et al, 2013;Kadriu et al, 2019).…”

Social crisis detection using Twitter based text mining-a machine learning approach

“…Using orange 3, an opensource data mining application [15] this study used the Liu-Hu sentiment analysis method for the following reasons: i) it provides a reasonable accuracy [16] with the dataset of this work; ii) Liu-Hu method in orange is free; iii) Liu-Hu in orange is easy to use for sentiment analysis; iv) it does not need a training dataset to analyze sentiment; and v) proffer a clear view of negative and positive sentiment based on the specific topic. Bag of words model in Spyder is used to classify text and to count word occurrence from social media and to identify the frequently discussed topic for the following reasons: i) bag-of-word is simple to understand; ii) easy to implement; iii) bag-of-word works faster for our dataset; and iv) many scholars prefer bag-of-word for [17] text classification.…”