PARAMETRICAL WORDS IN THE SENTIMENT LEXICON

In this paper, the main features of parametrical words within a sentiment lexicon are determined. The data for the research are client reviews in the Russian language taken from the bank client rating; the domain under study is bank service quality. The lexicon structure and the fragments from the lexicon database are presented. The sentiment lexicon includes two major classes (positive and negative words) and three minor classes (increments, polarity modifiers, and polarity anti-modifiers). This lexicon is used as the main tool for the sentiment analysis carried out by two methods: the Naïve Bayes and the REGEX algorithms. Parametrical words are referred to as the words denoting the value of some domain-specific parameter, e.g. a battery life, or time of waiting. To distinguish the main features of parametrical words, the parameters relevant for the bank service quality domain are determined. The results of the research demonstrate that parametrical words can be ranged neither in the positive class, nor in the negative one. The words denoting the increase of a parameter should be ranged in the increment class, as they intensify positive or negative emotions. The words denoting the decrease of a parameter should be ranged in a new class which may be called the decrement class, as they reduce positive or negative emotions. The revised lexicon structure including the decrement class is proposed. The evident progress on the way to the lexicon universalization can be achieved by distinguishing two special classes for lexical increments and decrements. Another helpful idea is to extract bigrams or trigrams which could include parametrical words and the domain attributes they refer to.