The present technology such as privacy-preserving data mining generates data, which is inherently uncertain in nature. There are other existing tools, which are also collecting data in an imprecise way. Further mining frequent patterns from uncertain textual data is not as simple as in precise data, and normal approaches that work for precise data are not applicable for uncertain data. This paper describes the motivation behind proposed method based on review of existing frequent termset mining techniques in document data. Further, a new mining method using karnaugh map is proposed for finding frequent termset from uncertain textual data, and experiment carried out requires only a single database scan for mining frequent patterns, which reduces to low processing time. © Copyright 2016 Inderscience Enterprises Ltd.