Artificial Intelligence & Big Data

FIT4BA research and innovation centre

Künstliche Intelligenz & Big Data 2

Artificial intelligence (AI)

While there are myriad formal and informal descriptions of intelligence, it is difficult to find a definition that gains universal scientific acceptance. Two example definitions are provided here; the first definition centres on thought processes and conclusions, whereas the second definition focuses on the meaningful and logically correct behaviour of an agent, such as a machine, vis-à-vis changes in a system.

The study of the computations that make it possible to perceive, reason, and act.

Winston, 1992

Computational Intelligence is the study of the design of intelligent agents.

Poole et al., 1998

Artificial intelligence (AI) thus comprises a multitude of technologies, tools, processes and associated process models. They are diverse as well as complex and can be combined to represent intelligent capabilities.

How can we determine whether a system is intelligent? Is there a difference between the intelligence of humans and that of computers? Can a machine think? A famous test designed to answer these questions is the Turing test, named after its creator Alan Turing. To pass this test, a machine must have the following capabilities:

  • It must be able to communicate in a natural language such as English.
  • It must have knowledge and be able to store it somewhere.
  • It must be able to reason based on this knowledge.
  • It must be able to learn from its environment.

There is no universally valid definition of artificial intelligence. However, there are conditions that a system must meet in order to be classed as intelligent. These include:

Systems that think like humans

In order to determine whether a system is cognitive, we first need to understand mechanisms that underlie human cognition. These mechanisms in the human brain can be described in different ways:

  • Introspection - observing our own thoughts
  • Psychological experiments - watching a person act
  • Neuroimaging - observing a brain in action

As soon as a sufficiently precise model of the functioning of the brain is available, it will be possible to reproduce this model using a computer program. The input and output of the program will then be observed. If the results of the experiment roughly approximate those obtained by humans, we could call the machines intelligent.

Systems that act like humans

The Turing test tests whether a machine can behave like a human being. The test evaluates results and actions that result from "intelligent" machine processes. An extension of the method originally developed by Alan Turing is the so-called Total Turing test, which also tests the perceptual abilities of machines by presenting them with physical objects, among other things. To pass this test, the machine must also have visual and robotic skills in addition to the four skills mentioned above.

Systems that think rationally

The Greek philosopher Aristotle was one of the first who attempted to establish what it means to "reason correctly" in the sense of it being an irrefutable argument. His syllogism provided templates for the structure of an argument. These structures always produced correct conclusions provided the premises were correct. An example would be "Socrates is a man. All men are mortal. Therefore, Socrates is mortal". These templates formed the beginnings of the field of logic. In the 19th century, logical notations were developed for all kinds of propositions. 1965 saw the first programs able in principle to solve all solvable problems written in logical notation. However, this approach had two problems. Firstly, it is not easy to render informal knowledge using logical notation. Secondly, there is a big difference between solving a problem in theory and solving it in practice.

Systems that act rationally

An agent is something that acts. It is expected to have capabilities such as working anonymously, perceiving its environment, persisting over a long period of time, adapting to change, and creating and pursuing goals. A rational agent acts such that it achieves the best result or the best expected result. All skills needed to pass the Turing test also allow an agent to act rationally.

Big Data

The amount of data generated globally is growing exponentially and continuously. The term "big data" refers both to the process of data generation and to extremely large data pools. One definition of big data involves the following properties:

  • Extensive: Huge amounts of data are not only generated by people, such as via Facebook and Instagram, but also by companies, for example by machines.
  • Diverse structure: Different types of data are stored, including unstructured data (see diagram).
  • Increasing speed: The rate of data generation continues to grow - and data volumes have to be stored and processed.
  • Unknown quality: The reliability of the data varies, as they often come from several sources (data may be uncertain/inaccurate).

For this reason, these data cannot be stored and analysed in conventional computer systems. They are too extensive, can change quickly and have a complex structure. The role of technology is to manage this flood of data. Efficient storage, distribution and dissemination of data is particularly important.

Big data combines information from many different sources. This includes web pages and social media or data from various sensors. Analysing these data, i.e. recognising trends and patterns, requires fast computers. Big data technologies are designed to create added value for companies, since the results can be used for statistics, surveys and other purposes.


  • Fasel, D., & Meier, A. (2016). Big Data: Grundlagen, Systeme und Nutzungspotenziale. Wiesbaden: Springer Fachmedien.
  • Gesellschaft für Informatik. (2013, Juli). Retrieved Jänner 9, 2020, from Big Data:
  • Poole, D., Mackworth, A. K., and Goebel, R. (1998). Computational intelligence: A logical approach. Oxford University Press.Russell, S., & Norvig, P. (2004). Künstliche Intelligenz : ein moderner Ansatz, 2.Auflage. München: Pearson Studium.
  • Villars, R. L., Olofson, C. W., & Eastwood, M. (2011). Big Data: What It Is and Why You Should Care. Framingham, Massachusetts, USA.
  • Winston, P. H. (1992). Artificial Intelligence (Third edition). Addison-Wesley.