In Big Data Analytics, a lot of raw data is available and abundance data collection exists. There are several mechanisms to collect data such as sensor based data collection where unique data will be provided.
For providing unique data using certain mechnisms, maintenance is required which are generally expensive. Computation of data item prices need to be calculated. This is one of the limitation in collecting data. To avoid such limitations, there are approaches auch as application allocation in E-ARL (Early Accurate Result Library). Revenue balance avoids starvation of mobile peers and E-Top incentive system which answers top-k queries in M-P2P networks by stimulating an economic market.
Data quality plays an important role. For good data quality there are key contributions from collaborative and non-collaborative methods. Here comes another limitation, the open research issue. An example for this limitation is the noisy data in photos which doesn't help in car plate recognition of traffic violations. Also, preserving data privacy is necessary to sustain long term data collection. Considering all this, incentives for crowd driven data collection can be linked. Data is obtained from crowd workers when incentives are given, the data obtained must be made into useful information.
However, the limitation incrowd sourcing is data reliability. Crowd sourcing and cloud serving are customer oriented where crowd sourcing is a potential labour.
No comments:
Post a Comment