Text Mining
According to a Forrester Research study, about 85% of a company’s business information is “trapped” in unstructured text documents (emails, internal reports, etc.). In other words, the vast majority of potentially useful information for a company is, in the current state, largely unknown to the company itself, external to the productive process. Evidently this is a very similar problem to that which, for structured data, the more consolidated Data Mining processes respond to. This simple example illustrates how useful extraction processes from texts via Text Mining techniques of new knowledge can be.
The step from text data, which is unstructured, to structured data, which is measurable, is not painless. Also, the usefulness of Text Knowledge Discovery systems is strongly conditioned by the quantity and quality of the knowledge that the available technology is able to extract from the text. The difficulty of this task has traditionally been a not insignificant hurdle to the achievement of Text Mining systems. But, in the recent past, the development of Natural Language Processing technologies and Database Knowledge Discovery, in addition to specialised software tools, have rendered this ambitious objective feasible.
Indeed, the technology which is currently available has allowed the implementation of solutions to real company problems. These solutions have shown to be successful case studies.