Back to Top

A Cognitive Map-Based Representation for Consumer Behaviour Modelling

Homenda, W., & Jastrzebska, A. (2017, November). A Cognitive Map-Based Representation for Consumer Behaviour Modelling. In the Eleventh International Conference on Advances in Semantic Processing (pp. 1-7). IARIA.
In many areas of science, there is a need for modelling approaches that represent knowledge in an abstract, generalized way. This need manifests itself mainly, when models are designed to be interpreted and used by human beings, but not exclusively. Abstraction comes in hand, when we deal with very large data sets. When standard numerical methods are inadequate to describe the data, it might be beneficial to turn towards granular models, based on concepts, where knowledge is aggregated and represented in an abstract fashion. Concepts-based methods facilitate smooth human-computer interactions as they allow to represent knowledge in form of relationships between phenomena – just like humans do.

An example of a concepts-based approach is a Cognitive Map and its generalizations. Cognitive Maps are described with weighted directed graphs. Nodes represent concepts (in other words: variables, phenomena, knowledge aggregates). Edges represent relationships between the nodes. Each edge has a weight quantifying strength of the relationship. Depending on a particular variant of the Cognitive Map, we can use different formalisms to represent concepts and relationships between them. If the model is crisp (as it is in the regular Cognitive Map model), nodes are sets and an element either belongs to a set, or it does not. Relationships between such nodes are represented as numbers from the set {-1,1}. If a model is fuzzy (Fuzzy Cognitive Map), nodes are realized with fuzzy sets. Then, an element belongs to a certain fuzzy set with a membership measured as a real value from the interval [0,1]. Relationships in the Fuzzy Cognitive Map model are expressed with real numbers from the [−1, 1] interval. Construction of a Cognitive Map consists of two phases: extraction of concepts and data-driven weight matrix learning. Cognitive Maps model causal relationships, and therefore, they can be applied to model phenomena that change in time. The idea, how to apply a Cognitive Map to represent time-dependent data is depicted in the Figure above.

In our research, we presented an application of the Cognitive Map model to represent consumer preferences. We showed and application of the Fuzzy Cognitive Map model to represent and illustrate changes in the demand for public transportation in the City of Warsaw. There were three major premises motivating the application of Fuzzy Cognitive Maps. Firstly, Cognitive Maps provide a stable behaviour irrespective of the length of the input time series. Secondly, the graph-based representation interface is human-friendly and provides an intuitive description of phenomena. Lastly, the concept-based representation will be beneficial in later phases of our investigations, in which we plan to join numerical data with text data describing conditions relevant to the public transportation system.


Human behaviour modelling is a prominent example of an area in which information processing and understanding remains a challenge. Among crucial problems of this domain are heterogeneity, multiplicity and uncertainty of information. Behaviour modelling benefits from methods that enhance understanding of dependencies between phenomena and form a comprehensive model over a collection of elementary information granules. If we consider analytical application, quantitative predictive modelling becomes obsolete, because it is unable to represent wealth of information and its structuring. Hence, there is a need for semantic knowledge modelling. In light of the above, we present a Cognitive Map-based modelling framework capable to represent decision making processes. The model assumes that motivational stimuli determine decision making outcome. In case of human decision making, needs play the role of motivational stimuli. A decision is an outcome of processing of human needs. In order to reflect this using a Cognitive Map-based model, we assume that concepts making a map correspond to various needs. In the paper, we present a processing scenario that applies a Cognitive Map of needs and a current state of personal stimuli to produce a decision. We also apply the model to real-world data in an experiment of mobile phone activity monitoring.
Bibtex Entry:
@INPROCEEDINGS {Homenda2017,
            author = {Homenda, Wladyslaw and Jastrzebska, Agnieszka},
            editor = {Homenda, Wladyslaw and Roman, Dumitru},
            title = {A Cognitive Map-Based Representation for Consumer Behaviour Modelling},
            booktitle = {SEMAPRO 2017, Proc. of the Eleventh International Conference on Advances in Semantic Processing},
            pages = {1--7},
            isbn = {978-1-61208-600-2},
            year = {2017}