Using speech to text technology, the goal of the project is to gather an extensive logbook of dialogues used by visitors during the 48 Stunden Neukölln main group exhibition. Simple telephones with headsets will record every word spoken by its user and is transmitted automatically to an online databank where it will be transcribed and analyzed using speech-pattern recognition software in real time.

The transcribed text is saved and stored, and subsequently run through an automated insight algorithm. This procedure will organize the text in such way that patterns of emotion  can be found, structuring the text as to give an overall idea of the perceptions of people visiting the exhibition. This data can be outputted again to project a computerized vision of the art on display. 

Only a select amount of visitors will be equipped with the technology. After each visit the devices are given to a new visitor. In this way, three to four people will be registered at any given time. Users are informed in advance of the functionalities of the system and are made aware of the possible breaches on their privacy, even though the set-up would produce entirely anonymous data. They can later view their own log, but are not allowed to make any changes. At the end, the entire log might be made public through different media. 

The project uses a corporate software which is integrated, one way or another, in almost every mayor Silicon Valley company (and beyond). Every call center, customer support, survey agency and peer service enterprise revolves around automated processes through SMS, telephone, VoIP or email communication. The gathering of data is their core business: gaining insight in consumer desires, anticipating demand, optimizing service structures, streamlining sales. Blindly trusting the word of Big Data, entire models of industry are based on the computationally generated analytics. 

Exhibition Insight Analysis therefor applies this logic on the art world, a scene of human interaction which is generally less prone to data analysis, and tries to find patterns of acceptation, rejections, disgust, euphoria, scepsis or other. The completely arbitrary reality in which the program will be forced to operate, the exhibition as opposite to the clear framework of consumer logic, will push the system to its limits. The generated output might become an aleatory collection of nearly, maybe, but not really.