Knowledge discovery through creating formal contexts

Knowledge discovery is important for systems that have computational intelligence in helping them learn and adapt to changing environments. By representing, in a formal way, the context in which an intelligent system operates, it is possible to discover knowledge through an emerging data technology called formal concept analysis (FCA). This paper describes a tool called FcaBedrock that converts data into formal contexts for FCA. This paper describes how, through a process of guided automation, data preparation techniques such as attribute exclusion and value restriction allow data to be interpreted to meet the requirements of the analysis. Examples are given of how formal contexts can be created using FcaBedrock and then analysed for knowledge discovery, using real datasets. Creating formal contexts using FcaBedrock is shown to be straightforward and versatile. Large datasets are easily converted into a standard FCA format.