Growing Networks – Modelling the Growth of Word Association Networks for Hungarian and English

In the new era of information and communication technology, the representation of information is of increasing importance. Knowing how words are connected to each other in the mind and what processes facilitate the creation of connections could result in better optimized applications, e.g. in computer aided education or in search engines. This paper models the growth process of a word association database with an algorithm. We present the network structure of word associations for an agglutinative language and compare it with the network of English word associations. Using the real-world data so obtained, we create a model that reproduces the main features of the observed growth process and show the evolution of the network. The model describes the growth of the word association data as a mixture of a topic based process and a random process. The model makes it possible to gain insight into the overall processes which are responsible for creating an interconnected mental lexicon.