“…In this section, the performance of our method is compared with other well‐known or recently proposed methods. Comparison of the proposed method was made against the following methods: (a) DPSO‐EDASum (optimization approach based on discrete PSO and EDA; Alguliev, Aliguliyev, & Mehdiyev, ), (b) LexRank (graph‐based approach; Erkan & Radev, ), (c) CollabSum (clustering and graph‐ranking based approach; Wan et al, ), (d) UnifiedRank (graph‐based approach; Wan, ), (e) 0–1 non‐linear (binary optimization based on discrete PSO approach; Alguliev, Aliguliyev, & Isazade, ), (f) QCS (machine learning approach based on hidden Markov model; Dunlavy et al, ), (g) SVM (algebraic approach; Yeh et al, ), (h) FEOM (fuzzy evolutionary approach; Song et al, ), (i) CRF (machine learning approach based on CRF; Shen et al, ), (j) MA‐SingleDocSum (metaheuristic approach based on genetic operators and guided local search; Mendoza et al, ), (k) NetSum (machine learning approach based on neural nets; Svore et al, ), (l) manifold ranking (probabilistic approach using greedy algorithm; Wan et al, ), (m) ESDS‐GHS‐GLO (binary optimization based on the global‐best harmony search heuristic, a greedy local search algorithm; Mendoza et al, ), and (n) DE (clustering and metaheuristic based approach; Aliguliyev, ). These methods have been chosen for comparison because they have achieved the best results on the DUC2001 and DUC2002 data sets.…”