Almost ten years ago, my prediction was, in next ten years neuro dynamic programming language will receive more adoption and become a common computer programming language. Variance in my prediction and actual result is significant. Recent announcement of unmanned car from google is very encouraging that my predication will become a reality one day.
In simple words, in machine learning field, there are two types of learning. Supervised learning and un-supervised learning. Unsupervised learning is a type of learning which makes a system to learn in dynamically changing environment. Neuro dynamic programming language concept is based on unsupervised learning type. Whereas, supervised learning are widely used in voice recognition (in your voice phone directory at office), face recognition and etc. Techniques like neural network, Bayesian learning are used in supervised learning.
Q-Learning, Dynamical programming are some of the techniques available in unsupervised learning. It is a method used by humans to learn and the framework is very simple. In a dynamic environment, the sequence of events are random and an action for the random events is taken. A feedback (reward) for the action is received and based on the immediate and long time reward of the action taken for the random event, a weight is assigned to the action. Based on the exploration and exploitation strategy of the system, weight assigned to the action in the past for the random event, the action is repeated if the same or similar random event happens.
In the current computation paradigm, programming logic is deterministic. In the future, deterministic logic is not sufficient in computation. A car that uses set of techniques to drive itself will be used to learn about that specific car. At any given time, for a given VIN, all the necessary details about the car will be available. It applies to all entities including humans.
Pagerank is an algorithm used by google to assign importance of each page in world wide web. The order of search results depends on the page rank assigned to each page. A web site has multiple pages and page rank assigned to each page in the web site. The page rank of the page depends on number of external links coming to the page and number of links going out of the page. The page rank of an incoming link also plays a role on determining the page rank of the page.
Page rank is used to decide the sequence of search results for a google search. Companies want their page to be at the top of the search result. The easiest way to be at top of the search result is to pay google for the adwords. The sponsored web sites appear above and to the right hand side of its regular search results.
Not all companies can afford to pay google for their Adwords. There is cost effective alternative which requires an understanding of how google works. Google’s internal working knowledge will aid to define analytical (or mathematical) strategies and implemented in the web site to improve the page rank and eventually with more visitors to the web site.
To validate the theory with emprical analysis, a plan with following steps drafted.
- Find a key word that is not in google’s index server
- Create a graph (random graph) and find the initial transition probability
- Evaluate the steady state of transition probability matrix using power methods – Which is the page rank of the graph (I will post the technical/mathematical details in a pdf. It is time-consuming to write matrix and other math notation in word press editor)
- Develop a set of web sites (pages) adhering to the random graph and each web site to contain the new key word
- Allow google crawler to include the new sites in their index server for the new key words
- Search for the new key word and absorb the order of search results
- Report the results
1. A new key word was selected and is given below. There is no google search result for the key word.
2. The random graph (a representation of how web sites are linked to each other) and the graph will be implemented with various blog post and each blog post will have the key word adhering to the graph. Each node of the graph denote a blog posting. The links connecting the nodes are the hyperlinks connecting the nodes.
Note: This page is used for google’s page rank emprical analysis. The links will be created based on the random graph created. This is node #1 which has the key word: xysivabodzinyx , xysivabodzinxy . As per the graph, it links out to page 2, page 4
The next generation computation devices are lurking around in TED as sixth sense devices and labs are experimenting contact lens devices to present the most relevant information in real-time with out manually seeking. The devices will search for relevancy and the present the information to the user. It is very similar to Terminator movie (please watch video carefully at 2.13) where the aliens receive the most relevant information for the given circumstance. The future devices will make the information readily available based on our circumstance, situation, and mood. Well, it is not science fiction any more and it will soon become a reality.
For those future devices, which are currently in experimental labs, the key component is an information gateway. Information gateway will seek relevant information for each user based on location, mood, and circumstance. The information gateway are nothing but the next generation smart phones. These information gateway will replace personal usage of PCs and laptops.
Google strategy team stuck a good balance to compete in the current market with Apple iPhone, RIM’s blackberry and laid basic foundation for information gateway market. Google launched Nexus one as their phone product today to consumer using Android operating system. Nexus one provides easy integration for all social networking tools and techniques. There are more detail comparison done between Nexus one and iPhone and this article focus is to study Google’s strategy and it’s alignment for future technology evolution.
Google’s strategy to provide options to consumer to select the service providers invites more customer base. However, I’m not excited about its pricing strategy. The device cost around $520 per unit. The pricing strategy will not let current iPhone users to migrate to Nexus one and also blackberry users will not quickly migrate to Nexus one since it does not focus more on running business application (like VIN locator, inventory management and etc).
Nexus one is an another great thing for Google but the unit price needs to come down..
Fifteen years ago, seeking for useful information in a public domain was a challenge and where as today, we have the opposite problem. There are lots of information available on any topic.
The social networking sites enables the right information come to us instead of we seek for it. But still, filtering the information is a key challenge. We more often go to search engine like google for information and filtering techniques in the google helps us to receive the most relevant information quickly.
Recently, I wrote a job aid/reference on “How to google effectively” and there are few beneficiary of that document personally stopped by my desk at work and thanked me for that job aid. That was signficantly modified by other team members in the organization and published to the entire organization.
We decided that live demonstration on utilizing google search engine using the job aid would give more clarity on usage. A work shop was scheduled and opened up to entire organization with the limit of 40 users and it filled up almost immediately. I conducted the work shop to the signed audience from business and IT area on how to google effectively using google keywords like file type, site, define, “wiTHin qUoTeS”, and etc.
I took a simple scenario (#1), search for presentation material in educational institutes containing our company name. Run the scenario as a novice search engine users and showed the total number of relevant documents returned by google was around 1.2 million documents. Used various google key words and step wise demonstrated how the results can be narrowed down and eventually reached 9 most relevant documents. In scenario (#2), demonstrated how to search effectively to understand the basic definition of crowd sourcing, scenario (#3) was to find the patterns registered in core business. Had few more scenarios and explained, in high level, how google works using their pagerank algorithm and the required knowledge level for internet marketing group to increase pagerank of our web sites and eventually to increase the number of potential customer visits to our web sites.
A scenario (#n) given by an audience on how to search satellite images of a used car dealership in America. The scenario is so broad and the criteria required refinement before we jump into search of ocean of information. The criteria was narrowed down just to Michigan. The results returned as a map data (kml extension file) and these files read by free software like google earth. The kml extension files contained locations of all used car dealership in Michigan with the satellite images. Since I didn’t have google earth installed in my laptop , I could not demonstrated the results back to the audience. The purpose of the write-up is to display the results of the kml extension file and raise the question if google should consider to directly read the kml extension file in their google maps.
Genchi Genbutsu approach for enterprise architect is an apt approach in some cases and this approach is apt for validating the viability of gmail for enterprise use. As a cost saving opportunity, the enterprise email infrastructure is one of the frequently visited area for potential saving by practical enterprise architects. It is often studied by hiring external infrastructure architect specialist in the email infrastructure area to perform a business case and cost benefit analysis and internal enterprise architects lead that study by providing the necessary internal information. To perform the initial study, generally, it cost enterprise not a negligible money and time. For an organization of size 10,000 employees, migrating an email infrastructure from Lotus Notes or Outlook to new email infrastructure will take at least 5-6 months (my estimate before I did this project) and requires coordination and training. Before the trigger is pulled to migrate the email infrastructure for cost saving opportunities, the business case and benefit analysis should be strong.
Recently, I lead and managed a team to study the gmail for enterprise use by following Genchi Genbutsu approach and my results are summarized below.
Team Size: 3
Duration of the project: 6 hours
No. of email users:100
Cost: $5020 ($5000 is refundable if pilot results are concluded as not successful before 30 days)
Steps followed to establish the Google app’s gmail
- Tried to use the free gmail for 100 users. Selected 100 users in the organization and tried to create 100 free gmail accounts. Gmail has strong spam protection measures. It didn’t allow to create more than 20 users. gmail is performing lots of user traffic analysis on their side and predicting the creation of multiple users. Even though, we wanted to create valid gmail viability using 100 legitmate users, google has build intelligence to prevent us creating more ids. Tried quickly to flush the cache in the browser, changed the browser, changed the PCs, changed the sub net still did not work. Google must be analyzing the user traffic by looking into the public ip (ie. web proxy to reach internet cloud) of our network. No customer service or support available for gmail. All the support material are available in the site and forums. No one available to talk to regarding the problem.
- Decided to use the google apps gmail to validate it.
- Bought new domain from godaddy.com (cost around $10)
- Configured the email gate way of the domain in registrar (godaddy.com) site by creating the MX records and setting up the priority. Good documentation available in the google app web site.
- In the online, bought 10 ids from google apps (using google’s checkout)
- Used their mass upload of ids to create all 10 ids. CSV file contains the user name, user id, initial password and provision to reset the password after their first login.
- All ids were succcesfully created.
- Took one id and send and received email. MX record took around 30-40 minutes to complete the configuration. After 30-40 minutes after the ids were created, we were able to successfully send and receive emails.
- Used the same format and added additional 90 ids to the CSV file.
- Used google check out to buy additional 90 ids ($4500). Additional 90 ids were not immediately added to the account after successfull completion of the transaction. Waited 40 minutes and tried to reach google check out support. There is no support/help desk number available. No contact information for sales. I felt, may be, my credit card company is not authorizing the transaction since the amount is considerably large. Called the credit company’s support and found out the transaction didn’t reach the credit card company yet. Waited for additional 20 minutes and assumed google checkout is not authorizing consecutive legitamate transactions in a short time (less than 60 minutes)
- Meantime, sent emails to their corporate support email address, google apps support email address, google checkout support email address screaming for “HELP”
- Repeated step 3-4 for a different domain and bought 100 ids this time. Transaction was complete and ids were not added to the account.. Got stuck again!!!
- Got a call from Google’s check out support. Told them what I wanted and second transaction was cancelled and additional 90 ids were added.
- Deleted all 10 ids I used to perform end to end test
- Tried to upload the 100 ids at one time.. Failed!!! Tried different combination, removed special character in the password, removed white spaces and few others. Nothing worked and everything FAILED!!!
- Tried to upload ONLY new 90 ids at one time. Successful!!!
- Once the ids were deleted, it can not be created immediately. It will take 5-7 days to recreate the same id. Strange!!
- For those 10 ids, had a different format ‘_’ instead of ‘.’ after the first and last name of the gmail address
- Loaded the 10 ids and it worked.
- Personalized the first page with company name, logo and etc.
- It took less than 1 hour (once the ID’s are successfully bought) to create all 100 ids.
- Each id has 25GB mail box size
- Administrator can manage all users, implement security policy, configure mail relays easily using the gmail console, restrict the traffic and etc
- Quick training material was developed easily
- SSL enforcement can be centrally done by the administrator
- Calendar, docs sharing, chat are part of the gmail
- It is apt for a small size (less than 1000 employees) company
- Fast, inexpensive, sufficient email space
Is google’s gmail ready for enterprise (10,000+ employees) use?
Short answer: No
Fuzzy answer: Almost there but not quite yet
Detail Answer:Support is the big concern. I could not talk to any one whenever I had a problem. I understand their support strategy. Provide all necessary information and empower the user to resolve the issues by them self. That is good strategy from google stand point. Not a customer friendly strategy. I had legitimate support issues during the check out and recreating the deleted ids. I had to wait for their support teams mercy whenever I had problem. Google is over analyzing the traffic and stopping the legitimate traffic. I can’t make 10,000+ employees depend on this infrastructure, when there is a problem, we need to wait for mercy of their support team to call us. Here is the screen shot of the 404 error received during the pilot (removed the company logo!!)
Wait for Google’s revised support strategy or ask your 10000+ employees to use free gmail for one or two years for their personal use before you transfer the corporate email infrastructure infrastructure to google app’s gmail.
Delivery of Google’s innovation mission continues and recent launch of google’s browser is an another prime example of their commitment in the innovation. There are already lots of articles surrounding Google’s Chrome and even an article suggested that chrome captured 2.74% market share in a day. I do not believe it and did not do any factual data analysis to verify it. It could be accurate statement for a smaller population but the analysis must be skewed and do not present the whole true market share.
Like millions of the curious people, I down loaded the Chrome on the day it was launched and played around it.
My immediate end user reaction on Chrome:
- Very simple
- Like the logo (Google’s predominant color in the logo)
- Most visited page is already a window (same as alt-tab in windows) in my first page. Very user friendly
- Integrated search with URL edit box
- History of web search is integral of the first page
- Major plugins works
- Looked almost same as the first version of IE in windows 95, when I was extensively using Netscape browsers
- Looked like extension of the google main page.
My immediate technologist reaction on Chrome:
- Very light weight
- Will be easily portal to many platforms (including Mobile platform)
Strategist perspective of Chrome:
- Google is on the right strategic path on launching the Chrome as their browser
- It is a stepping stone for google to completely dominate the next generation web
- Stepping stone to eliminate the notion of personal computer
- Cycling back to the mainframe environment. Yes, Ramba 3270, a thin dumb client and sends every thing to the mainframe to process. Chrome will be next generation 3270 will be sending everything to google to process it
- Chrome will be running in mobile devices like blackberry and the enterprise email will be running in google (gmail) and you will be connected all the time. You are already connected all the time through some means. In the future, it all will be google platform.
- An another browser war started, this time, Microsoft will be tasting their own medicine
Economist perspective of Chrome:
- If this continues, google will control the government in 20 years.
- Some body like Microsoft should react faster to keep up the pace of google, else soon they will grow so big to become monopoly in the technology domain and at the same time make all of us addicted to their technology. (20 years from now, Hail Google!! Hail Google!! will be universal national anthem for every nation in the planet Earth)