The overload of information available on the Internet has made text mining and simplified news and articles browsing an increasingly important user concern and a prioritized research issue. The aim of our project is to build a light-weight and effective text mining and summarization engine for the financial domain. This engine should also be easily trainable and adaptable to other domains. This paper describes a robust trainable user-focused summarizer for the financial domain that is adapted from algorithms by Kupiec, Pedersen and Chen (KPC) [4] and Lee, Goh and Kong [5]. It employs an adapted feature set to improve the robustness of the algorithm and to incorporate domain specificity in the engine. Evaluation tests verified the improved performance of our user-focused, domain-oriented, corpus-based approach over domain-independent approaches.