[Retracted] XAI-Based Reinforcement Learning Approach for Text Summarization of Social IoT-Based Content

Table 1

Statistical information of CNN/daily mail and XSum datasets.

Data set	Number of documents			Average document length		New in the gold standard summary
	Train	Verify	Test	Number of words	Sentence length	The proportion of binary groups that appear/%
CNN	10266	1320	1193	860.5	43.98	62.9
Daily mail	206961	14148	12397	7653.33	39.33	62.16
XSum	224045	2332	15334	654.7	29.77	93.31