Semantic memory organization and retrieval is a cutting edge topic that is being studied from different fields such as Linguistics, Psychology, Computer Science and Neuroscience. The aim of this thesis is to improve the understanding of conceptual organization and retrieval by means of network theory and the use of semantic verbal fluency tests (animals) in an unsupervised fashion. Conceptual organization will be studied here as a complex network attached to a dual-mechanism of information retrieval, i.e. switching and clustering.The chapters are organized as follows: 1. An introduction to the concepts of human brain, memory and network theory. 2. A study of the frequency patterns obtained from the verbal fluency tests. 3. Development of a statistical method for the unsupervised generation of a conceptual network and the insilico evaluation of switching and clustering. Such evaluation together with the definition of accessibility and diffusivity measurements allowed the decoupling of switching and clustering functioning. 4. Study of switcher random walks (by means of finite Markov chains) as an exploration-propagation paradigm in a number of insilico network models. 5. Modelization of the switching-clustering retrieval on the conceptual network obtained in chapter 3. 6. Amodel of concept acquisition and semantic growth based on frequency of concepts. 7. Study of the lexical access impairment in three different neurodegenerative conditions: Multiple Sclerosis, Mild Cognitive Impairment and Alzheimer’s disease. 8. General conclusions and outlook of this work.