Last month I talked about the explosion of unstructured information in retail and therefore the new technologies to tame its growth. This month I would like to debate how you’ll get actionable data from a number of that content.
Unstructured data includes new technologies like blogs and Wikis. Many companies have embraced these sorts of communications but do not have an efficient thanks to comprehend the content during a way which will be actionable.
Blogs now number within the millions and consistent with the newest issue of Forbes are growing at 100,000 per day. Anyone can now create their own free blog on Google’s Blogger.
Blogs have both made companies and crippled companies. Many companies are starting their own blogs to counter blogs of dissatisfied customers. There are numerous blog search engines but comprehending the range of content is just too much even for the foremost sophisticated searches.
By the way, blogs are old hat now.
Wikis are the most recent sort of community interface. There are even blogs about Wikis. Wiki-wiki is Hawaiian for quick. A Wiki is sort of a blog on steroids that permits content to be cross referenced, organized and straightforward to follow a thread.
Visit WIKI and you will understand them better, and make an entry in any subject that you simply have an interest in. Wikipedia is an on-line free encyclopedia that anybody can edit or create new topics. Actually one blog features a wiki joke.
Retailers are beginning to use the Wiki technology for knowledge management. Buyers, merchants, store managers and vendors can use the Wiki to quickly input unstructured information like a customer comment a few product or service.
They will input information a few competitor, marketing promotions, new colors for fashion, what’s selling and what’s not. This information is often made instantly available to everyone within the company who needs it.
Still who has time to read through the growing terabytes of knowledge for relevant content? IBM research has developed a tool for visualizing such evolving documents and therefore the postings of multiple authors.
It is called History Flow Visualization.
History Flow Visualization Application may be a “tool for visualizing dynamic, evolving documents and therefore the interactions of multiple collaborating authors”. The appliance includes online help, also as a plug-in for retrieving the history of a given page from any MoinMoin “wiki.” (MoinMoin is a complicated, wiki engine.)
How to get actionable data: How does it work?
History Flow Visualization Application represents each document as a vertical line whose length corresponds to the length of the document. The technology then applies a typical “diff” algorithm to successive versions of a document, using periods and angle brackets to point changes.
This level of detail is effective for free-form prose. After matching passages of successive versions are identified, the matches are represented onscreen by a parallelogram connecting the acceptable sections of the document segments. Both segments and parallelograms are colored to point authorship.
History Flow Visualization Application has four main visualization modes that allow the user to grasp the flow of changes and better understand the underlying pattern within the content.
Other on How to get actionable data Includes the Followings:
This is the default mode and it shows all contributions from different authors, color-coding the text to point the author of every sentence.
Individual author view:
This mode highlights the contributions of one author and it depicts the persistence of those contributions over time.
Recent Changes View:
This mode highlights the new content in each version of the Wiki page, independent of authorship. This view allows us to ascertain what portions of the text are edited the foremost over time. As an example this view might point to an emerging trend.
This mode has no colors representing authorship; instead, the main target is on the persistence of various contributions. A gray scale gradient goes from white (brand-new contribution) to dark gray (very old contribution).
The patterns revealed by History Flow Visualization Application show such information as spacing by date; occurrences of vandalism; authorship; growth; and persistence. Also read: 3 Advantages of Using Electronic Checks for a Business
The history flow application charts the evolution of a document because it is edited by many of us employing a very simple visualization technique.
Imagine a scenario where several people will contribute to a Wiki page at different points in time. Everyone edits the page then saves their changes to what becomes the newest version of that page.
History flow connects text that has been kept an equivalent between consecutive versions; in other words, it connects corresponding segments on the lines representing versions.
Pieces of text that don’t have correspondence within the next (or previous) version aren’t connected and therefore the user sees a resulting “gap” within the visualization; this happens for deletions and insertions.
How to get actionable data: A simple example
Here’s an example of an easy page with just a couple of edits: the primary eight versions of the Wikipedia entry for IBM. The page has three named authors (listed at left), including a script which changed some formatting.
Each author is given a singular color. Several anonymous authors also made contributions; their insertions are shown in reminder gray. The green regions show the contributions of the initial author, Peter Winnberg, many of which persist throughout the versions shown.
Text that persists over time is darkened to point its age, so at the proper side of the diagram Peter Winnberg’s contributions have changed from bright green to dark green.
There are many existing methods for visualizing document revisions. Several popular source control systems include the potential to color-code changed regions in files, and to point out a side-by-side comparison of two files, graphically connecting matching sections.
Other methods use a thumbnail view of a program, with line-by-line coloring to point authorship or age; see for instance the work by Eick et al. on software visualization.
History flow diagrams have some visual similarity to Theme River ™ and to Inselberg’s parallel coordinates, but our method depicts a totally different sort of data.
As far as we all know the timeline visualization introduced here is new, but please allow us to know if you’re conversant in other work we should always cite.
History Flow is out there to the general public on IBM’s Alphaworks Website
History Flow was developed by Martin Wattenberg and Jonathan Feinberg for visualizing patterns in very complex and dynamic collaborative content.