EMEA A&C CCOE Partner Technology Cloud Engineering

Exploring data streams with Oracle Stream Explorer - part 2

Webcast: Oracle's Fast Data Solutions at Oracle OpenWorldIn part 1 of this blog entry we’ve seen how
you can easily create a twitter stream with Oracle Stream Explorer, an add-on for Oracle Complex Event Processing, and feed
data into it.

In this second part we’ll overview how the stream can be

an Exploration

An exploration is a real-time insight over
a data stream, which can be filtered and summarised over specific time window.
And best of all, an exploration is also a stream at the same time – it can
provide input to further explorations or provide output to targets (csv file, REST,

Let’s say we are interested in twits per
minute for each given keyword that we feed into our twitter stream.

We start by
creating a new exploration from scratch

In the create wizard, we’ll enter “Twits
per minute” as exploration name and select Twitter stream as source.

As we are interested in number of tweets
per each keyword, we’ll summarize the stream by counting number of texts per
each keyword.

To specify the time window over which we’d
like to summarize the data, we need to click the right clock icon on the right.
We enter 1 minutes for range and 1 second for frequency. What this means is that every second we’ll recalculate number of tweets per minute for
each keyword. The frequency depends on how fast your data is coming in (and
also how up to date output you need).

As we configure the exploration you’ll
start to notice live data in the chart and table. But we can do more!

Fluctuation analysis

Now that we got the data, we can analyse it on the fly. For the purpose of this blog, let's say we are interested in fluctuations of twits per minute for each keyword. We start by creating a new pattern fluctuation

Notice that pattern is just a specific type of exploration which implements one of the supported pattern recognitions. To configure fluctuation, we'll enter twits per minute as event stream, keyword as tracking field and count_of_text as value field

Give it some time and you'll see textual and graphical presentation of fluctuation of the data over time. To get fluctuation for each individual stream, we should implement another exploration with filter for a specific keyword between Twits per minute and pattern fluctuation, but we'll leave this exercise to the readers.

To learn more about stream explorer and its underlying technology Oracle Complex Event Processing, attend our free webcast scheduled for 23rd of April!

Jernej Kaše is a Fusion Middleware Specialist working closely with Oracle Partners
in the ECEMEA region to grow their business by leveraging Oracle technology.


Be the first to comment

Comments ( 0 )
Please enter your name.Please provide a valid email address.Please enter a comment.CAPTCHA challenge response provided was incorrect. Please try again.