A Survey: Computing Iceberg Queries

Abstract

Knowledge discovery, tasks are deals with huge no of records. Queries are needed to identify unique attributes values and their aggregate that is above a predefined threshold from this huge number of records. This type of queries is called iceberg queries. Iceberg queries requires huge amount of main memory and takes longer time to answer the query. As computer system has limited amount of main memory, the processing of iceberg queries is a challenging task. This paper discusses different methods that are in literature for processing iceberg queries , we also explore pros and cons of these methods and future scope.

Data Stream Summary in Big Data Context: Challenges and Opportunities

“…Immediately, it is necessary to apply the methods of signal theory to them to summarize such data [35]. Among these efficient and robust mathematical tools, we find histograms [25], [26], wavelet compression, Fourier transformations, discrete cosine transforms, curve segmentation, etc. [27], [36], [37].…”

“…Furthermore, they can be used in order to summarize data which come from streams. The literature offers different types of histograms like Histogrammes V-optimal [11], [12], Equi-Width Histograms, End-Biased Histograms) [25], [26] or Compressed histogram [27].…”