This research paper proposes a novel Neighbourhood Rough Set based approach for supervised Multi-document Text Summarization (MDTS) with analysis and impact on the summarization results for MDTS. Here, Rough Set based LERS algorithm is improved using Neighborhood Rough Set which is itself a novel combination called Neighborhood-LERS to be experimented for evaluations of efficacy and efficiency. In this paper, we shall apply and evaluate the proposed Neighborhood-LERS for Multi-document Summarization which here is proved experimentally to be superior to the base LERS technique for MDTS.