You are viewing ![[info]](http://l-stat.livejournal.com/img/userinfo.gif?v=92.1){kind=small}
pankaj's journal
You are viewing pankaj's journal
|
|||||||
 |
|
|
|||||
|
Pankaj Mehra  This morning, as the fragrance of orange blossoms lingered in our backyard, diffuse sunlight and a gentle breeze provided the backdrop for a perfect Spring day. Our tomato plants had been sitting on the potting table, ready to go in. Along with two flats of marigolds, a first for us in this house. As I worked away at the day's chores, I had a lot to feel good about. The first was Charu's visit to UCSD. When she entered high school four years ago, Class of 2012 seemed like a distant concept. Even as the dates for SAT, apps, notifications, and decisions have come and gone, the whole process seemed unreal. No more. She was away, attending Triton day on the beautiful La Jolla campus, all by herself. We celebrated a lot of her Spring birthdays in this backyard ... she has blossomed here (learning to swim, learning to bike, learning to jump rope, all here). Now she is going to college. And I feel the joy of seeing my child blossom into the well-balanced, fun-loving, intelligent, hardworking, responsible woman she has become. Thinking of blossoms, one of the best fruits of my teaching labor has been a student named Mohit Talwar. He took several courses from me during my tenure at IIT Delhi, worked with me on my projects, and generally kept in touch as he worked first at Microsoft and more recently at Facebook. He and his family will be visiting us this week. I am really looking forward to it. As with children well-raised, one feels great pride (whether earned or not) in the success of those whose lives one touches as a teacher. That brings my attention to Whodini, a company borne out of an idea four of us brainstormed one beautiful day in August less than two years ago. Our product has been hitting significant milestones almost weekly. Large customers, cool new features, and most importantly a stage of product completeness that leads to engaged and satisfied users. It has blossomed too. Spring 2012 feels very special indeed. Current mood:  pleased.
Expert-finder HAL A startup called Whodini opened this week. Its core technology is an artificial intelligence engine designed to find experts inside your own company, and also present your own expertise. The company's algorithms can scan through your professional activity and figure out what you're good at. It then compiles a profile that, with your approval, is made available to co-workers or to the public. Read more here: http://www.arnnet.com.au/article/40 Current mood: optimistic.
Beyond Search: Context-Aware Computing Final submissions due: 1 July 2011 Publication date: March/April 2012 See CFP here. Please email the guest editors a brief description of the article you plan to submit by 15 June 2011 Context is the unstated actor in human communications, actions, and situations. It makes our communication efficient, our commands actionable, and our situations understandable to the people, organizations, and devices that provide us with content or services. The increased embedding of technology into our personal and social environments drives a need for context-aware computing. Context-aware computing offers mobile Internet users an experience that goes beyond user-initiated search and location-based services. Context awareness sharpens relevance when responding to user-initiated actions (such as product search and support calls). It also enables proactive communications through analysis of a user’s behavior and environment, thereby forming the basis for key business imperatives targeting customer-engagement systems. Even greater opportunity arises from context use in systems that can make sense of and engage in customer dialogs and forums. This special issue seeks original articles that support and illustrate context use in creating enhanced user experiences. Sample topics include
Contact Guest Editors Pankaj Mehra and Daniel Tunkelang (ic2-2012@computer.org) All submissions must be original manuscripts of fewer than 5,000 words, focused on Internet technologies and implementations. All manuscripts are subject to peer review on both technical merit and relevance to IC’s international readership—primarily system and software design engineers. We do not accept white papers, and we discourage strictly theoretical or mathematical papers. To submit a manuscript, please log on to ScholarOne (https://mc.manuscriptcentral.com:443/ic-c I am planning a session about context-aware computing and welcome your interest and suggestions in the lead-up to this session. The Ecosystem of Context: How to Play Speaker: Pankaj Mehra Level: Advanced Search, which was supposed to have been a technology of last resort, has become the de facto tool of choice in the age of information. However, as technology pervades deeper into our lives, we must look beyond search, and instead look to model, capture and exploit user context. This session is 50-50 strategy and software design, with the latter part focusing mostly on context data handling. This article in St. Petersburg times about UN Development Program report on Russia demographics highlights challenges facing Russia. The brain drain continues as well. There is lots of diffusion taking place from China in eastern parts of Russia. All in all, this should be seen as a net positive for China, which will see its influence spread even more (not just Asia, UK, Canada, Western US, Japan, Australia, but now also all of Russia). This influence translates into expanded markets for Chinese goods. For instance, I was surprised to learn during a recent visit that the bodies of most new buses in St. Petersburg were built by Chinese manufacturers.
My favorite non-fiction work is a book called "Knowing and Guessing" written by Satosi Watanabe . There is cute little argument in this book that every researcher in machine learning and adaptive methods will do well to understand. It is called The Ugly Duckling Theorem. And it shows that if all possible 2^(2^n) Boolean functions on binary n-vectors are considered, then there is no way any pair of binary vectors can be considered more similar than any other pair--n-pixel binary images of ugly ducklings and swans are equally similar--throwing a major wrench into the possibility of clustering and classification. Of course, this does not mean we cannot limit our attention to certain features and thereafter do clustering and classification. In fact, this is exactly what we do all the time. But the theorem opens the way to measure bias of knowledge representations and feature spaces. I was at the MMDS workshop at Stanford recently and different papers on classification, clustering and DR chose to use different distance measures in setting up their optimization algorithms. I am sitting there wondering how much magic is in the algorithm and how much in the distance function. It would be good to characterize the bias of distance functions the same way we characterize the bias of knowledge representations. For many years, scientists in statistics and learning used mean squared distance measure across a variety of algorithms and data sets, limiting the bias only to the data dimensions on which the algorithm was applied. With the creativity of the last decade in switching to clever distance measures, one must be careful accepting the results unless they have been demonstrated on a wide variety of problems and datasets. Creative definition of distance functions (without explicit biasing) will otherwise become the new black art form in machine learning. I really enjoyed the talk by Charles Semonyi sharing his space tourism experience. Saw the $100 laptop PC, which is way cool. Learned about a program between MIT and schools around the world that allowed a bunch of teenagers to genetically engineer "eau d'E. coli." Saw Martha Stewart and Sergei Brin among the attendees. But the most fascinating discussions I had were with a kindred technical soul, Jeff Jonas, SRD founder and now an IBM Distinguished Engineer. We have such resonating views on complex real-time information processing and event analytics that the discussion could immediately get to the guts of query processing and routing architectures and their relationship to the ingest-side processing. Saw a couple of issues of Make magazine: this one I am going to become a lifetime subscriber to! More after tomorrow's day-long sessions. Current mood: pleased.
http://wiki.oreillynet.com/scifoo07/ind I was just going through the attendee list and noticed the excellent set of people I am going to run into for 3 whole days. I haven't prepared a topic yet but maybe we will talk about some work I have been doing in combinatorial methods of interconnect design, a monograph I am working on for the past 6 years in my copious free time. Current mood: optimistic.
 Industry 1: Managing Richly Connected Information Pankaj Mehra Current mood: pleased.
|
|||||||
|
|
|
|||||
