Basic Text Analytics in E-Mail & Other Applications

I have wondered why many of the web-based or desktop-based e-mail providers do not include some smartness in their software. Some basic if not simple (to implement) things are:

Trying to send a mail without an attachment when the mail text says that there is an attachment.

Extracting meeting schedule information – date, time, place or web conference info and subject of meeting, from the mail text.

GMail has introduced some smartness into its service for detecting #1. I am not sure about #2 above, whether it is implemented or not. May be some beta version of it is already there. But I got curious today on the pop-ups like the following that I got.

But it remains to be seen how good their implementation is and one can never get a perfect one!!

Categories: Text Analytics Tags: ,

A Report on MAC India 2009

November 3, 2009 Leave a comment

I had been to Mobile applications Conference India last week since I eager to check how much the enthusiasm is, for mobile apps development locally in India? There were around 800 people and there were also 15-20 exhibitors.  Sun Startup Essentials & Forum Nokia were present but we were missing an operator presence – the organizer SiliconIndia should make sure that they bring in an operator – they are many now – to hear their point-of-view. I am happy to see a ‘critical mass’ developing here in this area. There were:

  • Arvind Rao from OnMobile which is one of the leaders in this space currently.
  • Prakash Sayini from Forum Nokia.
  • Bakshish Dutta from Sun Startup Essentials.

There were other speakers from TCS, InMobi, TeleDNA, MChek, Navteq, Satnav, MapMyIndia, etc. The event was well organized by SiliconIndia. I have summarized the key points for start-ups in this space, as advised by the various speakers & panel discussions:

I have shared some of the keep points that were brought out by the speakers & also during the panel discussions:

  1. Indian consumers are value conscious and there are very few people using high-end phones. Understand the needs of the  target niche segment of users for your mobile app first – instead of just developing a ‘me too’ app.
  2. There is no one killer app, there are many killer apps for different user segments.
  3. Speech and not data is the lowest common denominator for most of the phones & users. Main reason why  OnMobile started to develop speech based apps first. Their techies had experience from Philips & Nuance projects in Infosys.
  4. Be patient – OnMobile was not successful immediately. At one point they were even thinking of closing the shop but fortunately they had VCs funds.
  5. Time-to-market is critical since lifecycle of apps is short. Also need to constantly think of improving the app. Developers & operators do not need to worry about micro-billing & micro-promotions since network-side platform provide those features.
  6. UI and UX are the two main factors of apps to create user ‘stickiness’. Arvind indicated the Phone Back-up product for cumbersome UI.
  7. User roles, identity & profiles is also important.
  8. ‘AppStore model’ has created a level-playing fields for developers. But whether we need to share revenue with operators even for apps which require the network only for download is to be questioned!
  9. Which kind of applications? Amit Mehta of TCS talked about ABCD – Astrology, Bollywood, Cricket & Devotional. LBS applications still hold some promise. Whether video-based apps hold any promise was a question from audience but it requires 2.5/3G  adoption which has not got much usage in India yet!
  10. Last but not least – give ESOPs to make the whole team part of the company growth.

Of course, there was a separate panel discussion on LBS – Navteq, Interchain Solutions, Satnav & mapmyindia executives. I was interested in LBS based apps for Elderly Caredomain back in my days at Inst. for Infocomm Research (I2R) in 2002.  BUT security & privacy are THE MOST important issues still to be addressed, this WIRED magazine article is an interesting read – covers both the good & the bad sides of LBS & also some incomplete solutions!!!

Categories: Mobile Apps

A.P.J.Abdul Kalam on Leadership

October 30, 2009 Leave a comment

A few days back I had a post on C.K.Prahalad’s views on Leadership. A few days back I read a article in the local India edition of Entrepreneur about Abdul Kalam’s views on a Leader, which are the following (extracted from here):

  1. Leader must have a vision.
  2. Leader must have a passion to transform the vision into action.
  3. The above two are same as knowing where to go & not lose the sheep. Of course, the execution part is equally important & difficult as the vision part.

  4. Leader must be able to travel into an unexplored path.
  5. Leader must have courage to take decisions.
  6. The above two are related, one needs to be courageous esp. when treading unchartered territory.

  7. Leader must know how to manage success and failure.
  8. Example that was quoted for this quality was: to take self-responsibility in the event of a failure and also to praise others for success and not oneself. I guess this comes from ones humility – same as follow and not be ahead sheep-dog rule.

  9. Leader should have Nobility in management.
  10. Noble meaning generous or magnanimous.

  11. Every action of the leader should be transparent.
  12. Leader must work with integrity and succeed with integrity.
  13. The above two are related, we know what happened to Satyam Computers – mainly because of not doing these two.

Categories: Leadership Tags:

Universal Charging Solution for Mobile – At Last!!!

October 26, 2009 2 comments

Snap of the different types of mobile phone adapters that are there with me:

Mobile Phone Adapters Zoo

Mobile Phone Adapters Zoo

One can see the different types of the – in shapes, sizes and weights – on both ends. Not only that, the recent one pump in more current to charge faster. These are not the only ones I have, but there are a few more which I didn’t include. Not all of them are mine, some are left by the guests who came to my house !!! I am not sure what to do with all these. These will be thing of the past because of the GSM UCS standard to which many operators and manufacturers have committed themselves:

“The group has set an ambitious target that by 2012 a universal charging solution (UCS) will be widely available in the market worldwide and will use Micro-USB as the common universal charging interface. The group agreed that by the 1st January 2012, the majority of all new mobile phone models available will support a universal charging connector and the majority of chargers shipped will meet the high efficiency targets set out by the OMTP (Open Mobile Terminal Platform), the industry body who developed the technical requirements behind UCS.”

But we have wait until 2012 for this! May be we will have solar cells packed in the display itself by that time !!

Categories: Smart Phones

C.K.Prahalad on Leadership Styles

October 23, 2009 2 comments

Today I got some time to look at C.K.Prahalad’s views on Leadership, since I have been wanting to read about his ideas on FBoP. But listening & viewing him was much more interesting! Leadership is about hope, change & future. Always analogies help people to understand better and he likens a leader to a sheepdog. The following are the three important points he mentions:

  • Bark & not bite: Consensus building and listening to dissent & being open to new ideas which are different & valuable. Listening to customer voices for co-creation.
  • Follow and not be ahead: Leading from behind with humility and not be arrogant.
  • Knowing where to go & not lose the sheep: Should have a point-of-view about the future and be able to articulate it.

Leaders of future should have moral authority & not hierarchical authority: Gandhi Vs General George Patten styles. Gandhi was never dogmatic, autocratic but willing to changing his views. To him, means are as imp. as ends or goals and had clear values. But Patten used hierarchical leadership as a way of demonstrating credibility.

Categories: Leadership

Some LinkedIn Peeves

October 13, 2009 1 comment

Here are some of my peeves about LinkedIn from what I have tried in the last few days.

  1. They allow only 160 chars in the invite mesage, so I am forced to invent my own short forms:  for e.g.,LUT-on-OFW for Let’s Talk on One of Week-ends.
  2. No URLs in invite messages, I wonder what resembles like a URL in following text: “Since now I have time,shall we start on a book writing project? I thought you had some topics in your mind.LUT-on-OOW.”

I tried the MS-Outlook export as CSV & import into Outlook, no problem with that! I can rest assured that I have my LinkedIn contacts on my netbook also !!

Categories: Web 2.0 Tags: ,

yUML – Diagramming with a Difference

I found the yUML online diagramming service while browsing the Atlassian Confluence Blogs.
Unlike the pioneer of diagramming web service providers such as Gliffy, yUML processes the diagram input as part of the URL itself and returns an image file.  I am not sure how many are going to like this given that the URL can become pretty long for large diagrams for which this service is not meant for I guess. Though making it as part of URL avoids having any kind of back-end database, it becomes cumbersome. They also have a stand-alone install to be used within firewalled intranet.

The name yUML confused me initially since it sounds similar to the yWorks Diagramming Tools vendor and it thought it is using their diagram layout engine inside. But I found out that it is using the open source Graphviz instead. The Graphviz syntax has too many graph, node and edge attributes to remember which is hidden to a large extent by the more intuitive yUML syntax. But one disadvantage of that is the lack of flexibility provided by the Graphviz options which is visible in the sample Wiki Conceptual Graph I generated below. It remains to be seen how well one can strike a trade-off between the two extremes given that one is diagramming in a specific context – examples: UML diagrams, Network Diagrams, Organization Charts, Business Process Flows, etc. One should be able to define a DSL for each of these ‘domains’ (or contexts) and hide the nitty-gritty of graphviz node, edge & graph attributes behind the DSL translation.

The PDF does not come out well though, they still have certain things to iron out since it is still in beta version. But I strongly advise that the creators of this tool change the name to something else, to avoid the confusion with yWorks products.

I am not sure where the DSL of yUML is documented – may be not much to document, but here is one more (at least that I know of) textual UML notation in TextUML which is more verbose compared to yUML’s DSL.

Categories: Diagramming