A cognitive map is originally a term from psychology describing the mental representation that people or animals form of their surroundings. In management and AI, it also refers to graphical models that show concepts as nodes and their cause-and-effect relationships as weighted arrows, for instance “fuzzy cognitive maps”. Such maps help structure expert knowledge and explore scenarios, such as how price changes affect demand and stock levels.
Knowledge engineering covers capturing, organising and preparing expert knowledge so that IT systems can use it. Originally it focused on rules for expert systems; today the priority is often to document scattered knowledge from manuals, process descriptions and people's heads so that AI agents can answer reliably from it. This involves clear structures, up-to-date sources, defined responsibilities and closing identified gaps.
In computer science, an ontology is a formal model of a subject area that clearly defines concepts, their properties and the relationships between them. This allows people and machines to interpret data consistently and draw conclusions from it. Ontologies underpin many knowledge graphs, semantic search systems and data integration projects. Example: a product ontology states that an “e-bike” is a “bicycle” and has a “battery”.
Causal inference covers methods for deriving cause-and-effect relationships from data rather than merely identifying correlations. It answers questions such as 'What would have happened if we had not taken this measure?'. Tools include randomised experiments such as A/B tests, causal graphs and statistical techniques for observational data. Example: checking whether a discount campaign actually increased sales or whether they would have risen anyway for seasonal reasons.
Bayesian networks are graphical models that represent variables as nodes and their probabilistic dependencies as directed arrows. Using Bayes' theorem, they show how new observations change the probability of other events. One example is diagnostic support: estimating the most likely cause from observed symptoms. Their strengths are transparency and the ability to handle uncertain or incomplete data.
