Semantics management for big networks

Abstract

We define "Big Networks" as those that generate big data and can benefit from big data management in their operations. Examples of big networks include the emerging Internet of things and social networks. A major challenge in big networks is storing, processing and accessing massive multidimensional data to extract useful information for more efficient and smarter networking operations. Dimension reduction, learning patterns and extracting semantics from big data would help in mitigating such challenge. We have proposes a network "memory" system, termed NetMem, with storage and recollection mechanisms to access and manage data semantics in the Internet. NetMem is inspired by functionalities of human memory for learning patterns from huge amounts of data. In this paper we refine NetMem design and explore hidden Markov models, latent dirichlet allocation, and simple statistical analysis-based techniques for semantic reasoning in NetMem. In addition, we utilize locality sensitive hashing for reducing dimensionality. Our simulation study demonstrates the benefits of NetMem and highlights the advantages and limitations of the aforementioned techniques both with and without dimensionality reduction.