From Big Data to Big Data Mining: Challenges, Issues, and Opportunities

Exploring complex and big data

“…Traditionally, the relatively homogeneous data sources were verified and integrated into well-structured logical forms. However, applications of big data often require integrating data from both traditional and non-traditional sources (Che et al, 2013). With inexpensive sensors, mobile devices, the Internet, and social collaboration technologies, data are now generated in numerous forms such as text, web data, tweets, images, audio, video, log files, and many more.…”

“…Data inconsistency and incompleteness, imprecision, ambiguity and vagueness of available descriptions, latency of getting access to important data elements-all these imperfections have been partly addressed in the machine learning and data mining (ML/DM) community. Nevertheless, together with uncertainty, they are inherent and more important in many aspects of big data, particularly when learning from the social media, text, multimedia, language translations, or summarizing human opinions (Che et al, 2013).…”

“…Early on, DM algorithms were typically applied to thoroughly pre-processed data, which were assumed to come from well-defined sources. In big data scenarios, data sources have less clear origins, are unverified, and are in general of lower quality than in traditional systems (Che et al, 2013). Therefore, it becomes important to be aware of the provenance, quality, and value of such data (cf.…”

“…For a comprehensive review of the above requirements, the reader is referred to Che et al (2013) as well as Japkowicz and Stefanowski (2016b) for a detailed discussion on the differences between the earlier ML algorithms and newer big data approaches.…”

AHP Model for the Big Data Analytics Platform Selection

“…Data integrity, access control and accountability must be supported during the whole big data life cycle. Chen et al (2014) and Che, Safran, and Peng (2013) reviewed of state of the art frameworks and platforms for processing and managing big data as well as the efforts expected on big data mining. Also Singh and Reddy (2014) provided an in-depth analysis of different platforms available for performing big data analytics and assessed the advantages and drawbacks of each of these platforms based on various metrics such as scalability, data I/O rate, fault tolerance, real-time processing, data size supported and iterative task support.…”