A Distributed Storage Cluster Based Multi-Modal Content Retrieval by Machine Learning

The multi modal data, the combined form of information with video, text, and audio, is at its of peak of usage as in the various domains such as education, business, and research due to availability of the higher-level infrastructure. The multi-modal data is generated by various content creators for various purposes and the gain in the volume of such data is clearly observable. Also, the significant growth in multi modal data storage, streaming and information management services are gaining popularity. Nonetheless, the primary challenge for such services are the information or content retrieval matched with the lower time complexity and higher accuracy. The primary challenges for these information management services are the highly distributed storage and computational architecture of these services and the extraction of the information from various storage clusters with a proper synchronization is always a challenge. In the recent times, a good number of research attempts can be seen to solve these problems. However, these attempts are criticized for lesser accuracy during the distributed content retrieval and for higher time complexity due to improper handling of the replicated data. The proposed algorithms produce nearly 99% accuracy during the content retrieval process and nearly 20% improvement on time complexity compared with the parallel best research outcomes.