Rhythm in Free-Verse (Hindi)

Rhythm in Mukt-Kavita (free verse) . An article in Kaavyaalaya (co-authored with Dr. Vinod Tewary) that illustrates the presence and desirability of rhythm even in free-verse. The software Geet Gatiroop has been used.

Basic Structure of Hindi Poetry
Part 3- Rhythm in Mukt-Kavita (free verse)

An article in Kaavyaalaya (co-authored with Dr. Vinod Tewary) that illustrates the presence and desirability of rhythm even in free-verse. The software Geet Gatiroop has been used.

Browser Market Share

Three approaches to visualizing market share data using browser statistics. Of these, the Streamgraph variant seems to be most appropriate for storytelling.

In this article, we see three different approaches to visualizing market share data, using web browser statistics over the last 10+ years. The data is from W3Schools.com based on their log files. Unfortunately, there is no data for the 1990s, the emergence of the Internet. A time very dear, significant and nostalgic for me (as would be for countless others I am sure). Else, you would have also seen the likes of Lynx and Mosaic too. Maybe W3Schools was not around then.Browser Market Share Visualization
Vis 1: Browser Market Share – Streamgraph variant (D3 & Inkscape)

The visualization looks somewhat similar to a bar chart, but is significantly different. Here Time is on the horizontal x-axis, and the varying thickness of the “bars” (or width of the stream) indicate the percentage of market share of each entity.

This is a variant of the Streamgraph / Theme River visualization technique. Normally, in this visualization technique the different entities are stacked over each other. I have deliberately kept them separate, treating each entity as a separate stream. This enables one to see the shifts in market share from one browser to another (highlighted via the blue arrows) more clearly.

Here are some observations from the visualization:

  • Internet Explorer is still around, but steadily loosing ground since the last 9 years.
  • Firefox has also been steadily loosing ground since mid 2009, but not as drastically as IE.
  • And now more than half of the web browsing is done using Chrome.
  • While all this is quite common knowledge, the small but steady user base of Opera for 10 years despite Chrome, is fascinating. (aside: My nephew is one of those loyal 1.6%.)
  • The distinct shift of market share from Netscape to Mozilla to Firefox is apparent. It is essentially about the same code base being made available in a new name. Here is the history of those early years.

More conventionally, this data could also have been visualized as a line chart.
Vis 2: Browser Market Share – Line Chart (Excel).

Here too the very complimentary rise and fall of IE vs Firefox and then the rise of Chrome vs the fall of Firefox (and IE) is clear. If you look carefully, the distinct shift of market shares in 2004 from Mozilla to Firefox is also apparent (highlighted via the red circle).

However, as is often the case with linear scales and considerable disparity in data, the smaller values are all pushed down and become unclear. Even otherwise, the individual trajectories of each entity is not as clear as Vis 1. The steady presence of Opera; the default Mac browser Safari; the change of hands between Netscape, AOL and Mozilla are not directly apparent, or even when Chrome came into the picture.

Considering that this is market share in percentages where all values add up to 100, a Stacked Area Chart nicely shows the part-whole relationship.
Browser Market Share - Stacked Area Chart
Vis 3: Browser Market Share – Stacked Graph (Tableau). Click image for interactive Tableau version.

The ordering of the entities on the stack is crucial. It significantly determines the inferences and impressions drawn from a Stacked Area Chart. The layout on the left is what the software created first by itself. The short-lived AOL is by itself on top. The tapering of IE is clear but even though Chrome is gaining space, it gives an impression as if it is declining because its border line (with green Firefox) is declining.

In the layout on the right, the entities are ordered in the same order as in Vis 1. Now the rise of Chrome is unambiguous. Also, the way Firefox takes off from Mozilla’s (purple) market share is clear. This was not apparent in the ordering on the left because Mozilla and Firefox were disconnected, one at the top of the stack, one at the bottom.

The small players do not get obliterated here as in a line chart. This seems to be a better option for analysis as the market share is clearly mapped to the scale. In contrast, in Vis 1 (the Streamgraph variant) the scale of 0-100% is not clear. In fact, in the Streamgraph variant where each entity is an independent stream, there is a band of 0-100% range for each entity – the maximum width any of the players can attain.

However, for telling the story of the evolution of browsers’ market share, Vis 1 seems more appropriate. The individual trajectories are clear. One can show the stream flow animated over time and introduce the relevant explanatory arrows and annotations in due course at appropriate junctures of the historical narrative. The metaphor of market share clout as a swelling or trickling stream is poetic. Animated Stacked chart emerging from left to right over time does not seem as intuitive.

What are your thoughts on the pros, cons, preferences of the options presented? What other appropriate ways of visualizing this data would you suggest?

Other related visualizations:

This visualization by Michael VanDaniker is very beautiful. However, by the author’s own admission it does not deliver much functional value. It was made primarily to demonstrate the use of the open source data visualization framework: Axiis. Click on the image for the full-interactive version. It is very nice.

This one is somewhat funky. It delivers almost 0 functional value, other than the small bar chart on the top left. The position of the browser-planets or the why the countries lie on an inner or outer orbit, seems totally arbitrary.

Software Launch: Visualizing Rhythm in Hindi Verse

We can sense the rhythm in verse while we read it. Poets and song writers can sense it while they write. How can we tangibly see the rhythm in written verse?

We can sense the rhythm in verse while we read it. Poets and song writers can sense it while they write. Those writers clued-in into the technicalities of meter in verse, know how to check their meter by counting the beats on their fingers, in their heads or by jotting down on paper.

Many techniques have been created to visualize the rhythm in music. How can we tangibly see the rhythm in written verse?

The software “Geet Gatiroop” is a visualization technique which shows the rhythmic patterns in Hindi poems and songs.

These lines of the famous poem Satpura Ke Ghane Jungle by Bhawaniprasad Misra are visualized as:
Satpura Ke Ghane Jungle - Viz

Without the text:
Satpura Ke Ghane Jungle - Viz without text

You can obtain the visualizations by putting in the lines of Hindi text in Devanagari Unicode in the software interface. At a basic minimum, it gives the maatraa count of each line of the poem/song. This is the first and most basic aspect of rhythm in Hindi verse. Maatraas are the units that determine meter in Hindi verse, just as syllables do in English verse.

However, the software also shows many other significant patterns, such as the maatraa alignment, rhymes and other patterns and embellishments (alankaar). The visual mapping of the shapes and colors bring out the phonetic patterns.

The software can be used by:

  • Poets and Song Writers: For refining the structure of a poem or song while writing. Song writers can also use it to write verse as per a rhythmic structure given by the music composer.
  • Music Composers: For what-if analysis to see how they can fit a verse into the rhythm of their music. They can make modifications such as line breaks, adjustment of maatraa allocation and pronunciation, combining of letters to obtain the structure they need.
  • Literature and Linguistic Experts/Enthusiasts: For structural analysis of existing works of Hindi verse.

To get you started, three videos are available which introduce the concepts of maatraa counting using the software.

Part 1: Short Long Vowels / Laghu Deergh Swar (2:32)

Part 2: Flexibility in Urdu (6:56)

Part 3: Half Letters / Aadhaa Akshar (10:27)

More content will come your way in the following weeks.

I invite you to come, try out the software here – with your own writings or those that you love.

Resume Visualization

A simple way (using just MS Excel) to create a crisp resume that displays your history on a timeline. A KISS design that facilitates comprehension without excessive visual excitation.

A company asked me recently to send them my updated resume.

I enjoy data visualization. Also, I am now trying to simplify my life towards essentials.
So this is what I came up with –

Click on the image for the full hyper-linked PDF version.

I like it for the following reasons:

  • It is one page and elegant.
  • The main information contained in any resume (the education and employment history) is presented as a data visualization. The information is arranged in a manner which visually facilitates the easy assimilation of the information, apart from explicitly reading the text. There are specific reasons why I like this data visualization which I will discuss below.
  • There is room for customization [such as the interests and skills], to tailor the content as per the intended audience.
  • It seems to be more in line with my personality, specially compared to my earlier run-of-the-mill resume.

    Now, a bit more regarding the visualization.

    Various forms of resume visualizations are there on the internet today. Below are some thumbnails. Click on them for the full view of these and other examples.

    In comparison to whichever resume visualization that I have seen on the internet so far, and independent of them too – these are the specific reasons why I like the one that I cooked up:

    • Bereft of excessive visual stimuli, it follows the KISS principle. Thus the objective of a more effective communication does not lose out against the agenda to impress. Infact it is so simple that it may hardly seem like a data visualization, but it is. The layout in fact is essentially the same as the left side of the first thumbnail [if you click it and see its full view]. However, it seems to me that, being visually simpler and more in line with conventional writing of text, makes this layout easier to assimilate.
    • The spatial and color coded arrangement of information has the following impact
      • gaps in education or employment are highlighted
      • relative time spent at each engagement is more clearly apparent.
      • provides space to write about each engagement in keeping with the amount of time spent at that engagement. Thus it encourages one to keep things brief.
      • enables one to see simultaneous engagements. Normally these are listed separately in resumes.
    • Gaps in history and relative time spent in engagements are derived information from resumes that HR specially looks out for. I suppose people in HR become adept at mentally identifying these when they glance through resumes. Such a layout would facilitate things for them.
    • The information that the visualization highlights would be largely apparent even in smaller thumbnail-type versions. Two examples are given below.
    • Lastly, it is made using MS Excel. So anyone can make a similar one using any spreadsheet software. No special visualization tool is required.

    The layout can be tailored to specific individual trajectories and on the choices made of what one wants to highlight. Here are some possibilities:

    • Many people first finish their official education which is followed only by work life. In such a case, a long empty left column may be a waste of space. One can either then keep the learning and working entries in the same column, with different color coding, or use the learning [education] column to mention other professional trainings, certifications, workshops during one’s work life, that the individual may want to highlight.
    • Two ways come to mind for including awards and recognitions.
      • By including an icon [like a star] and relevant text, in the box of the relevant engagement. In this case however, there may not be enough space for mentioning the award/recognition completely, or against the correct year.
      • If learning and working are merged into one column, with different color coding, then space is available for a separate awards and recognition column.

    I offer this layout to you dear reader, in case you want to use in your own resume or in your organization.