Databases Reference
In-Depth Information
•
Data Workload-1:
Streaming Analytics.
This type of workload
essentially consists of processing streaming data with predefined
behavioral patterns at real time. Once a pattern is observed then
real time responsesare formulated.
•
Data Workload-2:
High Velocity Data Ingestion.
There are
several subpatterns in this type of workload.
•
You can simply keep collecting the data without applying
any transformations; this data at a later point in time can be
analyzed. The intent is to not to lose the data streams as they
happen.
•
In other scenarios, you may have to collect the data,
transform and analyze all at the same time to contextualize
the data.
•
Data Workload-3
:
Linkage Analysis.
Primarily these types of
workloads are meant to establish relationships and linkages
between different states of data. This workload is computation
and read intensive as node statistics need to be computed and
children of a node need to be read dynamically.
•
Data Workload 4
:
Rare-Event Detection.
Looking for a specific
pattern from the vast data sets across multiple attributes is a very
dataanalysis workload.
•
Data Workload 5:
Data Mash-Ups.
Usually in these types of
workloads you are developing a story line or creating a “data bag”
linking not only data attributes but also events that happened
in isolation and may not have significance. But taken together
as a string of events occurring in a timeline, their importance
amplifies especially across multiple event streams. Sequence
analysis linking pieces of events together are some of the common
examples of these types of workloads.
•
Data Workload 6
:
Text Analytics.
A very commonly observed
data workload in big data scenarios: sentiment analysis, opinion
mining, social network analysis etc., fall majorly in this category.
•
Data Workload-7
:
Time Series Analysis.
This type of work loads
deals with pattern detections and occurrence or non-occurrence
of specific events across moving time windows of data.
•
Data Workload-8
:
Data Forensic.
This workload is primarily
triggered by data scientists exploring large data sets with questions
previously not thought of. They cast a wide net and often come up
with few patterns. The query patterns in this type of data workload
are both “depth search” as well as “breadth search.”
