Compared　with　general　group　decision　making,　it　is　more　difficult　to　reach　consensus　on　emergency　decision　making　(EDM)　due　to　its　complexity　and　urgency.　Experts　from　a　variety　of　backgrounds　and　levels　of　knowledge　can　collaborate　to　solve　a　particular　problem,　and　for　this　reason,　linguistic　terms　are　often　utilized　to　express　experts＇　opinions　because　of　their　flexibility　and　ease　of　use.　However,　linguistic　expressions　are　emotional　and,　therefore,　more　difficult　to　reach　a　consensus.　Taking　into　account　the　inherent　emotional　characteristics　of　linguistic　expression,　a　three-way　semantic　scales　model　with　emotional　preference　is　proposed.　Existing　minimum　cost　consensus　(MCC)　models　are　not　suitable　for　this　situation　because　the　cost　of　adjustment　in　various　semantic　categories　is　different　and　asymmetric.　Therefore,　in　the　context　of　EDM,　a　minimum　cost　consensus　model　(MCCM)　with　linguistic　information　is　proposed,　considering　three-way　semantic　scales　and　asymmetric　adjustment　costs.　First,　a　novel　MCCM　with　semantic　category　constraints　(MCCM-SCC)　is　proposed,　which　considers　the　asymmetric　adjustment　cost.　Then,　taking　into　account　the　degree　of　tolerance　of　experts　in　semantic　difference,　an　extended　MCCM-SCC　with　semantic　difference　is　developed,　and　related　theorems　are　proved.　In　addition,　a　comprehensive　feedback　mechanism　is　developed　considering　the　regret　aversion　of　decision　makers　to　prevent　the　manipulative　behavior　of　experts.　Finally,　a　case　study　is　used　to　further　elaborate　on　the　methods　and　models　mentioned　above　and　highlight　their　rationality　and　effectiveness　through　comparative　analysis.