The dataset is a CoNLL 2003 NER dataset that is composed of sentences with wordsper  line  5  possible  labels  :   “LOC,  ORG,  MISC,  PER,  O”,  that  represent  the  NERclasses.   The  “LOC,  ORG,  MISC,  PER,  O”  labels  are  abbreviation  of  “LOCATION,ORGANIZAT ̇ION, MISCELLANEOUS, PERSON, OTHER” respectively.  T