An Efficient Approach for News Summarization using Map-Reduce and Clustering

The ability to effectively summarize news articles is crucial in effectively managing the vast amount of available information. This research presents a novel approach to news summarizing, utilizing the combined functionalities of MapReduce and clustering methodologies. The limitations of traditional methods become apparent when faced with the vast and varied nature of news material, resulting in less-than-optimal summarizing results. The methodology employed in this study involves utilization of the MapReduce architecture for the simultaneous processing of extensive textual data. During the Map phase, summaries are generated in an independent manner by dividing news articles into smaller, more manageable segments. The final phase of clustering improves the process of summarization by grouping sentences that are semantically similar inside each chunk. This innovative methodology guarantees not only the capacity for expansion but also an enhanced capability to encompass a wide range of elements present in the original news content. The effectiveness of the suggested MapReduce and Clustering Summarization methodology is demonstrated through experiments conducted on CNN-daily dataset. Our research findings demonstrate enhancements in the quality of summarization, the speed of processing, and the capacity to manage larger datasets in comparison to traditional approaches. We were successfully able to make use of ideas of distributed computing and clustering algorithms which allowed us to generate news summaries that are all-inclusive as well as relevant in their contextual understanding. Our main aim of this research is to dig deep through the application of Map Reduce and clustering methodologies in field of news summarization within a research context.