Updates from baid RSS Toggle Comment Threads | Keyboard Shortcuts

  • baid 11:09 pm on March 28, 2012 Permalink | Reply  

    Confluence of mobile sensing and social sciences 

    I was recently drawn towards the growing use of mobile phone sensors and online social networks in understanding social behaviors. The Human Dynamics Group at MIT Media Labs seems to have a number of interesting projects on this topic. The following papers are quite related to things we discussed in the class and are very interesting to read:

    Social fMRI: Investigating and shaping social mechanisms in the real world

    Describes a mobile-phone-based social and behavioral sensing system and some initial results on how to use them to study the connection between individuals’ social behavior and their financial status, network effects in decision making. The implementation of mobile sensing is of particular interest: they develop a very easy to use open source framework called funf with which app developers can easily develop context sensing apps. The results are mostly preliminary: One shows that there is correlation between the common apps downloaded and the people you hang out with.

    How Many Makes a Crowd? On the Evolution of Learning as a Factor of Community Coverage

    This paper uses the framework from the paper above to constructed a SMS messages social network using info about more than 97,000 SMS messages sent. The aim is to predict each participants’ personal information by using their friends personal information. The results look quite impressive; the prediction accuracy for many attributes like religion, ethnicity, origin, age, etc. seems to be above 0.95. The title of the paper comes from their study on how the prediction accuracy is increased when the users sample set grows.

    VN:F [1.9.22_1171]
    Rating: 5.0/5 (3 votes cast)
     

    • anu 6:08 am on October 19, 2012 Permalink | Reply

      gud

      VA:F [1.9.22_1171]
      Rating: 5.0/5 (1 vote cast)

    • coupon codes 11:51 pm on October 30, 2012 Permalink | Reply

      I just got engaged a week ago and am looking into Confluence Park and Franklin Park conservatory. I wanted to know if anyone has any advice or personal stories about these locations.

      VA:F [1.9.22_1171]
      Rating: 5.0/5 (1 vote cast)

  • baid 7:15 pm on March 25, 2012 Permalink | Reply  

    Quantifying Facebook friendships 

    Since Facebook expanded its set of developer APIs, a number of small and big companies are trying to make sense of the wide range of data available. A key component of the exposed information is of course the social graph with people as nodes and facebook friendship as the edges. The harder question is how to deduce which of your friends are closest to you and might influence your preferences or decisions. That’s the aim of an interesting and well-cited (considering just 2 years since publication) paper which appeared in WWW’ 2010: Modeling Relationship Strength in Online Social Networks

    The key technique used is actually not surprising: Its the familiar latent variable model from Q4 in the course. The paper hypothesizes that “In online social networks therefore, we can model the relationship strength as a hidden effect of nodal profile similarities. Such profile attributes include, for instance, the schools and companies the users attended, the online groups that they joined, the geographic locations that they belong to, etc.”

    The paper explains the design of the inference of relationship strengths from the list of interactions between them and shows how it substantially  improves the correlation for a few different classification problems.

    Applications: From the list of citations, it seems like such a latent variable model can readily be used for social influence, suggesting more friends, risks and fraud detection and recommendation systems.

     

     

    VN:F [1.9.22_1171]
    Rating: 5.0/5 (4 votes cast)
     

  • baid 8:47 pm on March 14, 2012 Permalink | Reply  

    How about operator agnostic mobile services? 

    In the context of the class discussion on Q11 and Q12, here’s an interesting operation/business model that was proposed in a Mobicom’11 paper:

    Let’s assume our future phones to contain multiple 3G/4G network interfaces that can be simultaneously used with each interface potentially connected to a different provider. A new entity will act as a ‘service aggregator’ and the user will only have a contract/contact with this agency. On the back-end, the service aggregator will have tie-ups with several mobile service providers and will connect each mobile device with the ‘most appropriate’ provider at a fine time granularity.

    Now there seem to be two models that can be used, the first being described in the paper and the second being mentioned as unfeasible:

    1. Each provider broadcasts the set of parameters that are needed to make the decision on which operator to select, viz the current load and the price. Based on this info, each mobile device, on its own, selects the ‘best operator’ based on its location and application requirements.
    2. The service aggregator can potentially collect all the information from multiple providers and make this decision centrally. The tradeoff seems to be computational complexity vs. performance gain here.

    In either case, it seems like an interesting new operation model worth exploring more. Whether or not the operators would like such a model remains to be seen.

     

    VN:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
     

  • baid 8:57 pm on March 8, 2012 Permalink | Reply  

    Telcos: Transport service providers or more ? 

    A fair share of the blogs and news articles covering the recently concluded Mobile World Congress had a lot to say about the evolving role of the telecom providers in the mobile ecosystem.

    The general grievance from the provider side is summed up by this comment by Sunil Mittal, chief executive of Indian carrier Airtel: “We’ve become the bad gatekeepers. When somebody watches YouTube on a mobile phone and ends up [with a] big bill, he curses under his breath at the telecom operators. But YouTube is consuming a massive amount of resources on our network. Somebody’s got to pay for that.”  The same thought is shared by a number of other telecom CEOs blaming both the consumers and the content providers for stretching their networks.

    However true this problem might be, it has attracted a lot of flak by angry consumers (who happen to write a lot of blogs). This one for example, blames the telcos themselves for being in such a crunched situation. The author argues that the cellular companies are hardly doing anything to align their interests with those of the customers or of the content providers; hence this collision course is inevitable. I think that the whole market is in such a flux in the present times that its probably understandable if the telcos take a fair amount of time to incorporate any major changes in their business model. However, a number of companies are starting to see the synergy required and are making efforts towards tie-ups with content providers.  This article, for instance, points out that there is an ever increasing support for VoIP, IPTV, cloud services and WiFi access by telecom providers; and this shows their understanding of the consumer demands.

    VN:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
     

  • baid 6:56 am on February 29, 2012 Permalink | Reply  

    TED Talk on why videos go viral 

    Just found this funny and interesting talk about viral videos from Kevin Allocca, a trends manager at YouTube: http://www.ted.com/talks/kevin_allocca_why_videos_go_viral.html

    He describes three key reasons behind viral videos: Tastemakers, Community participation and Unexpectedness.

    The tastemaker angle is essentially what we studied in the class: nondescript videos gaining massive viewership after an influential person tweets or blogs about it or mentions it in his TV show. The example given is that of the ‘double rainbow‘ video which went on to get more than 30 million views after Jimmy Kimmel tweeted about it. Here’s the current snapshot of its viewing history. We can clearly see the point at which that tweet occurred.

     

    The other two factors: Community participation and unexpectedness are also clearly exemplified in the talk but seem to be very hard to build a theory around. My take is that unexpectedness is probably too subjective and content centric to be amenable for maths but some kind of analysis can be devised for community participation.

    VN:F [1.9.22_1171]
    Rating: 0.0/5 (0 votes cast)
     

  • baid 3:33 pm on February 20, 2012 Permalink | Reply  

    Startups in the TV White Space domain 

    A note on the burst of activities we saw after the TV White Space ruling by FCC: Seems like a number of companies were formed in the 2007-2010 period, though activities seem to have slowed down over the last 12 months or so. Opportunity for innovative wireless technology or another new tech bill killed by the politics: remains to be seen !

    While a host of big companies are already involved in various aspects of TV White Space (see a recent survey on TV White Space Cognitive Radio Patent portfolio); a stream of startups are emerging with an eye on the TVWS applications. Brief info on some of them:

    1. Spectrum Bridge: Formed in 2007, is leading the field trials for TVWS deployment and is one of the Database Admins assigned by FCC. Focus: database, application software, value added services, and full network deployment
    2. Adaptrum: Small Silicon Valley Startup, formed in 2004 with ties with UC Berkeley. Built one of the 1st White Space hardware for FCC approval along with the big players – Microsoft, Philips and Motorola. Focus seems to be hardware implementation but they also mention real-time resource monitoring, automated resource management, and self forming and optimizing networking capability in their company profile.
    3. Neul: UK based startup founded in 2010. Aims to use the TV White Space for machine-to-machine communication like smart-grid, asset tracking, vehicular systems, remote health monitoring etc. Team looks impressive, not much impact as yet.
    4. Key Bridge Global: A 2001 company which joined the TVWS bandwagon as a database administrator, though the revenue model is unclear for all the database providers. Have created a subsidiary called dsa, with some generic details about TVWS on the website for now.
    A long list of interested parties can be obtained from the current members list of Wireless Innovation
    VN:F [1.9.22_1171]
    Rating: 5.0/5 (1 vote cast)
     

c
compose new post
j
next post/next comment
k
previous post/previous comment
r
reply
e
edit
o
show/hide comments
t
go to top
l
go to login
h
show/hide help
shift + esc
cancel