Alcatel-Lucent is in a data-centric business. Equipment and systems are built to transport valuable data streams generated by network service subscribers. The operations of these systems also produce large volumes of usage and performance data that worth exploration. Novel ways for inference with data can uncover the knowledge contained in these massive data streams and archives, and provide many opportunities for enhancing the systems and services. Discovery of hidden correlations between

more »

... rk health and network state indicators can help prevent failures and improve system design. Swift, accurate characterization of network usage allows the service providers to personalize their offerings, detect malicious traffic, guarantee quality of service, and optimize resource allocation. For business operations, in-depth analysis of marketing data, service logs, and logistic records can provide valuable insight. Intelligent processing of the data being transported on the networks can lead to innovative new applications and value-added services, and generally, extend the functionalities of communication networks to reach new market segments. We discuss example uses of automated inference with data along these themes, and highlight the challenges and opportunities to be addressed by continuous research. Tin Kam Ho (Contact Author) Bell Labs, Alcatel-Lucent Room 2C-381,