Created for my thesis project, E.V.G is an interactive text analysis tool that allows users to view the data hidden within works of literature through the application of generative design. The project was presented as an interactive installation at the Institute of Art Design & Technology Dunlaoghiare where user could use the application with the help of custom motion tracking software.
EVG presents the user with a stepped process by which a piece of chosen text is converted into data and used to generate a unique nodal system, allowing the user to explore and interact along the way.
This process encourgaes the user to consider the changing form of the data. From text to visuals to an interactive system.
Extract
Once a text has been chosen by the user they progress on to the extraction stage. In this stage the user witnesses the process by which data is extracted from their chosen text. First, the text is rendered so that the user may read it if they choose to. The words composing the text are then colour coded based on their parts of speech (noun, verb, pronoun, etc). The colours used are generated by the program based of the characteristics of the text. Based on the parts of speech tags, lexically irrelevant words are removed from the text leaving behind only lexically relevant parts of the text. A lexically relevant word is a word who meaning is important in conveying the overall meaning of the text. Generally nouns, verbs and adjectives are considered lexically relevant. Once only the lexically relevant words remain they are sorted based on their part of speech tag and numerical data extracted from the text is displayed to the user.
Visualise
In the next stage the data extracted is visualised in an interactive, animated infographic. The occurrence of every lexical word in the test is represented along a radial distribution. Each concentric ring represents a different part of speech group (noun, ver, pronoun, etc) and each stroke within that ring represents an instance of that particular part of speech.
The occurrence rate of lexical words in the text is visualised using a bubble map that surrounds the radial distribution. Each bubble corresponds to a word that appears in the text. The size of each bubble corresponds the the number of times that word occurs within the text.
Generate
The final stage in the program is the generation of a nodal system. The generation of the system is seeded using the data extracted from the text. Attributes such as sentence length and syllabic complexity among others influence the colour, physics and and behaviour of the system resulting in a unique generation for each unique piece of text.
The central node corresponds to the main body of the text. The child nodes correspond to the paragraphs that compose the main body and the grandchild nodes correspond to the sentences that compose each paragraph. This nested structure, when visualised, results in the nodal system.
