Table 7.

Proposed algorithm comparison with similar algorithms (advantages and disadvantages).

Criteria	Algorithms
	SARPPIC	COV-1 and COV-2
Recommendation entities	Profile integration of social properties, i.e., BC and tie strength/social ties as entities for recommendation which is very appropriate for COVID-19 contact tracing.	These algorithms do not utilize social properties, i.e., BC and tie strength as entities for recommendation.
Cold-start and data sparsity challenges	Reduction of cold-start and data sparsity challenges due to its (SARPPIC's) capability of utilizing social properties, i.e., tie strength/social ties (through contact durations and frequencies) and BC (through shortest paths).	These algorithms utilize traditional collaborative filtering (CF) methods as entities and therefore the effect of cold start and data sparsity is not as minimal as compared to that of SARPPIC due to less social property inclusion.
Algorithm performance in terms of evaluation metrics	In terms of utilized evaluation metrics, namely, precision, recall, F1, and AM (Tables 5 and 6), SARPPIC outperforms COV-1 and COV-2 in relation to effective generation of people-to-people recommendations (COVID-19 patients) due to robustness, suitability, and effective social property inclusion for efficient contact tracing.	In terms of utilized evaluation metrics, namely, precision, recall, F1, and AM (Tables 5 and 6), COV-1 and COV-2 do not perform to the level of SARPPIC in relation to people-to-people recommendations (COVID-19 patients) due to nonutilization of social properties for efficient contact tracing.