In　the　recent　years,　the　social　foraging　behavior　of　E.　coli　bacteria　in　human　intestine　has　been　used　to　solve　optimization　problems.　This　paper　presents　an　improved　K-Means　algorithm　involving　bacterial　foraging　in　order　to　overcome　the　defects　in　K-Means　algorithm.　The　paper　applied　both　the　new　clustering　algorithm　named　K-Means-BFA-K-Means　algorithm　(KBFAK)　and　the　elementary　K-Means　algorithm　to　the　fabric　data　items.　It　is　found　that　the　KBFAK　has　the　same　results　as　K-Means,　but　its　execution　time　has　been　largely　reduced.　©　2012　Praise　Worthy　Prize　S.r.l.　-　All　rights　reserved.