GLOSSARY
What does the term AI mean?
We explain the most important terms and technologies related to AI.
D
Deductive Intelligence
...is a new type of technology developed by INFORM for the WorkforcePlus workforce management system. It makes it possible to separate the technical and logic levels during development. The modular logic is expressed in derivation rules, so-called deduction rules. This allows the system to make deductions on its own without having to describe in the rules where and how they are to be applied. This enables simple and fast implementation of individual customer requirements. The deduction rules are executed on the DI Engine.
F
Fuzzy Logic
…mimics the human way of communicating, modeling, and reasoning. It rarely uses terms that are only true or false. While in classical logic, a statement can only be either true or false (0 or 1), the truth value of statements here can also assume intermediate values (e.g. 0.2 = not very true or 0.9 = fairly true). In this way, decision rules can be formulated, such as “If the containers are quite light and the container stack is quite close to the quay wall, then do not stack very high there.” (…to minimize risk in the event of a storm).
H
Hybrid Artificial Intelligence
Under this term, we at INFORM integrate operations research and artificial intelligence, including fuzzy logic and machine learning. The combination of computer algorithms and human expertise leads to results that are far superior to traditional management and purely data-driven algorithms.
K
Artificial intelligence
…is a branch of computer science that deals with the study of mechanisms of intelligent human behavior. On the one hand, the ambitious goal is to imitate human behavior and thinking. On the other hand, it is about an automatic and autonomous execution of tasks within clearly defined areas. For this purpose, various methods are used, including neural networks, machine learning, heuristic methods, and fuzzy logic. In everyday life, AI can be found in speech and text recognition systems, among others. In medicine, AI supports doctors during operations and diagnoses. Companies use AI primarily to solve complex optimization problems.
M
Machine Learning
…is a subfield of artificial intelligence. By recognizing patterns in existing data sets, IT systems are able to independently find solutions to problems. The goal is to intelligently link the data, recognize correlations, draw conclusions, and make predictions. For this to succeed, the systems must first be trained with relevant data. Rules for analyzing the data set and recognizing patterns are also needed. Subsequently, machine learning systems can calculate predictions and probabilities for events, adapt to changes independently, and optimize processes on the basis of recognized patterns.
O
Operations Research
...is the technology on which all INFORM systems are based. It includes state-of-the-art methods of Operations Research (OR) and Computational Intelligence. OR can be used for planning as well as for the “real-time” control of time-critical processes. Companies use OR to solve, among other things, warehousing, sales, allocation and distribution problems, as well as problems in calculating optimal routes. This is often very complex: finding the shortest possible travel route through 15 cities for a sales representative theoretically leaves 43,589,145,600 possibilities. With OR, companies find the best possible decision alternative from such an unmanageably large number.