Тилини ривожлантириш департаменти ўзбекистон республикаси олий ва ўрта махсус таълим вазирлиги

Stylistic features.
The stylistic features of the text tend to describe it lexical and structural. 
Vocabulary richness measure, frequency of special characters, distribution of word length, letters n-
grams, special characters, digital n-grams, frequency of letters are used to determine the stylistic features 
[El-Masri, 2017: 5].
102
** Associate professor, Samarkand State University, [e-mail]  
103
***Assistant, Samarkand State University, [e-mail] 

---

Alisher Navoiy nomidagi Toshkent 
davlat o‘zbek tili va adabiyoti 
universiteti 
“O‘ZBEK MILLIY VA TA’LIMIY 
KORPUSLARINI YARATISHNING NAZARIY 
HAMDA AMALIY MASALALARI”
Xalqaro ilmiy-amaliy konferensiya
Vol. 1
№. 01 (2021) 
290 
Syntactic features.
Syntactic features describe the structure of sentences and parts of sentence and 
how words sequence are related. Syntactic features of the text are often used for opinion classification and 
sentiment analysis. These include n-grams of words, n-grams of POS tags, n-grams of lemmas, positive 
adjectives after nouns, and negative adjectives after nouns and others. 
Part-of-speech features.
POS tagging is the assignment of lexical categories to words in a text. 
That is, to determine whether words in a text belong to categories such as verbs, nouns, adjectives, 
adverbs, numbers, pronouns, auxiliaries, conjunctions. The POS tags represent the distribution of words 
in a specific order. This sequence can be used as a text feature. 
Lexicon-based features. 
Lexicon-based features are also used to increase the performance 
accuracy of the classification algorithm. Sentiment analysis dictionaries (sentiment lexicons) are used for 
this purpose. Usually, sentiment lexicon dictionaries present a word and its corresponding sentiment 
score, a value that helps to opinion classification in a text.