To avoid the complexity of natural language for computer programs, a proposed approach is suggested which simplifies the language into sequences of predefined keywords. The approach is: 1) the relevant proposed keywords should be obtained by applying patterns and rules in the text. 2) New keywords must be defined and the corresponding idea must be explained by those. 3) The statistics of the new keywords and patterns should be systematically collected. 4) Statistical analysis or data mining techniques will help in detecting sub-optimal situations. 5) Progressively this will lead to the generation of an optimal set of keywords and patterns and the final knowledge encoded will be easy to mine using simple programs. Read more at: http://www.datasciencecentral.com/profiles/blogs/feasibility-of-a-personal-knowledge-management-system-based-on