Both in the market- and industrial environment there is a need for certain knowledge and describing either the relative positions of two actors in the market or the relative position of a specific company to it's competitors.
With the spread of the World Wide Web several online news sources and sites became accessible, which provide these kind of information, although in an indirect way. Such natural language information are rich in the aforementioned knowledge answering these questions.
Int he first phase of my thesis I've collected and processed these news, in order to formalize them with text mining methods and build a model which can easily be understood with common sense. During the task I processed the collected news with the natural language processing method, to gain proper nouns. After filtering the results the components formed a graph. This later on visualized the results using a graph drawing algorithm.
These processes are separable operations on the various data sets, so connecting them leads to the development on softw
are architectural environment which generates a business connection graph from the news inputs.
To work up the processes I used Python and functions realizing the services mentioned beforehand.