id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
ajsts-4667	Catayoc, Rabel	Public Perceptions of Telegram’s Association with Illicit Activities: A Sentiment Analysis Using VADER and Machine Learning	2025	22	.pdf	application/pdf	13313	635	12	Computational linguists these results emphasize the necessity of building sentiment models sensitive to the nuances of negative expression in public discourse. Consequently, sentiment models should integrate more contextually sophisticated approaches, including neural network-based or transformer models, which are capable of better representing nuanced linguistic features such as irony, sarcasm, implicit negativity, and brevity that characterize authentic digital discourse.	cache/ajsts-4667.pdf	txt/ajsts-4667.txt