Cognitive Immune System-Based Co-Evolutionary Information Retrieval Model

The existent models of information retrieval (IR) focus on the relationship between index terms and documents. However, the current retrieval requirements demand delicate control to show the user preferences and feedback. And the exact results are emphasized for information explosion. Inspired by the maintenance function and model in immune system, a novel IR model is proposed based on cognitive immune system (CIS). Term network models the similarity relationship between terms and documents, the distance and similarity relationships among terms. The query string, user profile and feedback message stimulate the cytokine network which is a runtime structure to for IR. Co-evolutionary mechanisms are also designed to enable the emergence features of the model inspired by immune system.