Table 22.

Research topic prediction based on term frequency using the selected Weka algorithm.

Rank	Term	Observed	Term	Observed	Term	Predicted	Term	Observed
		2018		2019		2020		2020
1	DATASET	0.019411	Dataset	0.019293	Embedding	0.020756	Dataset	0.020833
2	EMBEDDING	0.012028	Embedding	0.018099	Dataset	0.017509	Embedding	0.015237
3	ANNOTATION	0.008888	Encoder	0.009572	Encoder	0.011884	BERT	0.01076
4	LSTM	0.008571	LSTM	0.008271	BERT	0.008609	Annotation	0.009168
5	DNN	0.006005	Decoder	0.007093	Decoder	0.008261	Encoder	0.009156
6	SR	0.005689	LM	0.006079	Classifier	0.007376	LM	0.006342
7	RNN	0.005585	Metric	0.005929	LM	0.006825	Transformer	0.006299
8	Encoder	0.005373	BERT	0.005745	Metric	0.006738	SR	0.006232
9	Classifier	0.005365	SR	0.005388	LSTM	0.006276	Metric	0.00604
10	Neural network	0.005334	Annotation	0.005326	Transformer	0.004887	LSTM	0.005866

Predictions are marked in green.