Table 4.

Trending topic classification approaches and findings

Authors	Year	Approach used	Findings
Zubiaga et al. [43]	2011	SVM with 15 different features	Classified current events with an accuracy of 82.9% memes with 73.1%
Lee et al. [20]	2011	- MNB with bag-of-words TF-IDF - C5.0 decision tree learner	MNB accomplished 70% and C5.0 decision tree 65% accuracy
Zhu [42]	2018	MNB with short text aggregation	Model achieved 73.33% of accuracy and build and classifies in 1.5 seconds
Shalini et al. [33]	2019	- Bag of Tricks classifier - CNN - Bi-LSTM	The best were Bag of Tricks, then slightly worse CNN and last Bi-LSTM
Liu et al. [21]	2019	CNN-LSTM (A mix of a CNN and a LSTM)	Classification binary of offensive tweets attained 98% of F1-score and 67.9% of F1-score on classified sentiments