Document Indexing Strategies in Big Data A Survey

Abstract

From past few years, the operations of the Internet have a significant growth and individuals, organizations were unaware of this data explosion. Because of the increasing quantity and diversity of digital documents available to end users, mechanism for their effective and efficient retrieval is given highest importance. One crucial aspect of this mechanism is indexing, which serves to allow documents to be located quickly. The problem is that users want to retrieve on the basis of context, and individual words provide unreliable evidence about the contextual topic or meaning of a document. Hence, the available solutions cannot meet the needs of the growing heterogeneous data in terms of processing. This results in inefficient information retrieval or search query results. The design of indexing strategies that can support this need is required. There are various indexing strategies which are utilized for solving Big Data management issues, and can also serve as a base for the design of more efficient indexing strategies. The aim is to explore document indexing strategy for Big Data manageability. The existing systems like, Latent Semantic Indexing , Inverted Indexing, Semantic indexing and Vector Space Model has their own challenges such as, Demands high computational performance, Consumes more memory Space, Longer data processing time, Limits the search space, will not produce the exact answer, Can present wrong answers due to synonyms and polysemy, approach makes use of formal ontology. This paper will describe and compare the various Indexing techniques and presents the characteristics and challenges involved.