Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

HarambeeNet: Data by the people, for the people

1,737 views

Published on

Published in: Technology, Education
  • How are you today i saw your profile in (slideshare.net)and i love it please i will like to be your friend .please you can write back to me in private email address (sussy02pet@yahoo.co.uk)so that it will enable me to give you my pictures.
    yours sussy.
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here
  • Be the first to like this

HarambeeNet: Data by the people, for the people

  1. 1. Data by the people, for the peoplePowering Interactions via the Social Web<br />Michael Bernsteinmitcsail | user interface design group | haystack group<br />mit human-computer interaction<br />
  2. 2. Computer Science<br />“In the most basic sense, a network is any collection of objects in which some pairs of these objects are connected by links.”<br />- Easley and Kleinberg, page 2<br />[Zachary ‘77, via Easley and Kleinberg ‘10]<br />
  3. 3. With the abstraction, we can: - Reason at high levels - Make predictions - Interact online - Model data<br />http://www.flickr.com/marc_smith<br />
  4. 4. Social Science<br />“The analysis of patterns of social relationship in the group is then conducted on the graph, which is merely a shorthand representation of the ethnographic data.”<br />- Zachary ‘77<br />[Zachary ‘77, via Easley and Kleinberg ‘10]<br />
  5. 5. Methodological mismatch<br />Many of you are sitting on terabytes of data about human interactions.  The opportunities to scrape data – or more politely, leverage APIs – are also unprecedented.  And folks are buzzing around wondering what they can do with all of the data they've got their hands on.  But in our obsession with Big Data, we've forgotten to ask some of the hard critical questions about what all this data means and how we should be engaging with it.<br />- danahboyd, WWW ‘10<br />
  6. 6. Methodological mismatch<br />Many of you are sitting on terabytes of data about human interactions.  The opportunities to scrape data – or more politely, leverage APIs – are also unprecedented.  And folks are buzzing around wondering what they can do with all of the data they've got their hands on.  But in our obsession with Big Data, we've forgotten to ask some of the hard critical questions about what all this data means and how we should be engaging with it.<br />- danahboyd, WWW ‘10<br />
  7. 7. building privacy-sensitive systems<br />building successful systems<br />
  8. 8. Netflix: Getting it right<br />Collaborative filtering<br />http://www.eecs.berkeley.edu/~zhanghao<br />
  9. 9. Netflix: Getting it right<br />Temporal dynamics<br />[Koren ’09]<br />
  10. 10. the challenge<br />bridging<br />
  11. 11. Soylent<br />UIST ‘10<br />Eddi<br />UIST ‘10<br />FeedMe<br />CHI ‘10<br />Collabio<br />UIST ‘09<br />
  12. 12. Soylent A Word Processor with a Crowd Inside<br />human computation<br />marketsvoting<br />[Bernstein et al. UIST ‘10]<br />
  13. 13. Interface<br />Wizard of Oz<br />
  14. 14. Highly-educated workers, mostly from the U.S. and India<br />Appropriate for generic cognition tasks with little intrinsic motivation<br />
  15. 15. Interface<br />Wizard of Turk<br />Wizard of Oz<br />Wire paid human computation directly into an interface<br />
  16. 16. Editing for length is excruciating<br />Even experts make writing mistakes<br />High-level decisions result in lots of small tasks<br />
  17. 17. Shortn: Text Shortening<br />
  18. 18. Blog – 83%<br />Print publishers are in a tizzy over Apple’s new iPad because they hope to finally be able to charge for their digital editions. But in order to get people to pay for their magazine and newspaper apps, they are going to have to offer something different that readers cannot get at the newsstand or on the open Web.<br />Classic uist– 87%<br />The metaDESK effort is part of the larger Tangible Bits project. The Tangible Bits vision paper, which introduced the metaDESKalong withand two companion platforms, the transBOARD and ambientROOM.<br />Draft uist– 90%<br />In this paper we argue that it is possible and desirable to combine the easy input affordances of text with the powerful retrieval and visualization capabilities of graphical applications. We present WenSo, a tool thatwhich uses lightweight text input to capture richly structured information for later retrieval and navigation in a graphical environment..<br />Rambling E-mail – 78%<br />A previous board member, Steve Burleigh, created our web site last year and gave me alot of ideas. For this year, I found a web site called eTeamZ that hosts web sites for sports groups. Check out our new page: […]<br />Technical Computer Science – 82%<br />Figure 3 shows the pseudocode that implements this design for Lookup. FAWN-DS extracts two fields from the 160-bit key: the i low order bits of the key(the index bits) and the next 15 low order bits (the key fragment).<br />
  19. 19. Crowdproof: Human Proofreading<br />Finds errors that AIs miss, explains the reason behind the problem in plain English, and suggests fixes<br />
  20. 20. The Human Macro<br />Macro scripting without programming<br />‘‘Please change text in document from past tense to present tense.’’ <br />I gave one final glance around before descending from the barrow. As I did so, my eye caught something […] <br />I give one final glance around before descending from the barrow. As I do so, my eye catches something […]<br />
  21. 21. The Human Macro<br />Macro scripting without programming<br />‘‘Pick out keywords from the paragrah like Yosemite, rock, half dome, park. Go to a site which hsa CC licensed images […]’’ <br />When I first visited Yosemite State Park in California, I was a boy. I was amazed by how big everything was […] <br />http://commons.wikimedia.org/wiki/File:03_yosemite_half_dome.jpg<br />
  22. 22. The Human Macro<br />Macro scripting without programming<br />‘‘Hi, please find the bibtex references for the 3 papers in brackets. You can located these by Google Scholar searches and clicking on bibtex.”<br />Duncan and Watts [Duncan and watts HCOMP 09 anchoring] found that Turkers will do more work when you pay more, but that the quality is no higher.<br />@conference { title={{Financial incentives […]}}, <br /> author={Mason, W. and Watts, D.J.}, <br />booktitle={HCOMP ‘09}, […]<br />}<br />
  23. 23. Programming Crowd Workers<br />Rule of Thumb: 30% of worker effort on open-ended tasks will have an error in it<br />Two useful personas: The Lazy Turker and The Eager Beaver<br />
  24. 24. The Lazy Turker<br />Does as little work as necessary to be paid<br />The theme of loneliness features throughout many scenes in Of Mice and Men and is often the dominant theme of sections during this story. This theme occurs during many circumstances but is not present from start to finish. In my mind for a theme to be pervasive is must be present during every element of the story. There are many themes that are present most of the way through such as sacrifice, friendship and comradship. But in my opinion there is only one theme that is present from beginning to end, this theme is pursuit of dreams. <br />
  25. 25. The Lazy Turker<br />Does as little work as necessary to be paid<br />The theme of loneliness features throughout many scenes in Of Mice and Men and is often the dominant theme of sections during this story. This theme occurs during many circumstances but is not present from start to finish. In my mind for a theme to be pervasive is must be present during every element of the story. There are many themes that are present most of the way through such as sacrifice, friendship and comradeship. But in my opinion there is only one theme that is present from beginning to end, this theme is pursuit of dreams. <br />
  26. 26. The Lazy Turker<br />Does as little work as necessary to be paid<br />The theme of loneliness features throughout many scenes in Of Mice and Men and is often the dominant theme of sections during this story. This theme occurs during many circumstances but is not present from start to finish. In my mind for a theme to be pervasive is must be present during every element of the story. There are many themes that are present most of the way through such as sacrifice, friendship and comradship. But in my opinion there is only one theme that is present from beginning to end, this theme is pursuit of dreams. <br />
  27. 27. The Eager Beaver<br />Go beyond task requirements to be helpful, but introduce errors in the process<br />The theme of loneliness features throughout many scenes in Of Mice and Men and is often the dominant theme of sections during this story. This theme occurs during many circumstances but is not present from start to finish. In my mind for a theme to be pervasive is must be present during every element of the story. There are many themes that are present most of the way through such as sacrifice, friendship and comradship. But in my opinion there is only one theme that is present from beginning to end, this theme is pursuit of dreams. <br />
  28. 28. The theme of loneliness features throughout many scenes in Of Mice and Men and is often the dominant theme of sections during this story. <br />This theme occurs during many circumstances but is not present from start to finish. <br />In my mind for a theme to be pervasive is must be present during every element of the story. <br />There are many themes that are present most of the way through such as sacrifice, friendship and comradeship. <br /> But in my opinion there is only one theme that is present from beginning to end, this theme is pursuit of dreams. <br />The Eager Beaver<br />Go beyond task requirements to be helpful, but introduce errors in the process<br />
  29. 29. Find-Fix-Verify<br />A design pattern that controls the efforts of the Lazy Turker and the Eager Beaver<br />Separates open-ended tasks into three stageswhere each worker makes a clear contribution<br />
  30. 30. Find<br />“Identify at least one area that can be shortened without changing the meaning of the paragraph.”<br />Independent voting to identify patches<br />Fix<br />“Edit the highlighted section to shorten its length without changing the meaning of the paragraph.”<br />Soylent, a prototype...<br />Randomize order of suggestions<br />Verify<br />“Choose at least one rewrite that has significant style errors in it. Choose at least one rewrite that significantly changes the meaning of the sentence.”<br />
  31. 31. Why Find-Fix-Verify?<br />Why split Find and Fix?<br /> Force Lazy Turkers to work on a problem of our choice<br /> Allows us to merge work completed in parallel<br />Why Add Verify?<br /> Quality raises when we put Turkers at odds with each other<br /> Trade off lag time with quality<br />
  32. 32. Data is made of people,<br />Data is made by people,<br />Data is made for people.<br />
  33. 33. Collaborators<br />Rob Miller, David Karger, Greg Little, Katrina Panovich, David Crowell<br />Mark Ackerman<br />Björn Hartmann<br />…and about 9000 Turkers.<br />I am generously kept off the streets by an NSF GRFP and NSF award IIS-0712793. <br />
  34. 34. Blog<br />Print publishers are in a tizzy over Apple’s new iPad because they hope to finally be able to charge for their digital editions. But in order to get people to pay for their magazine and newspaper apps, they are going to have to offer something different that readers cannot get at the newsstand or on the open Web.<br />Classic uist<br />The metaDESK effort is part of the larger Tangible Bits project. The Tangible Bits vision paper introduced the metaDESK along with two companion platforms, the transBOARD and ambientROOM.<br />Draft uist<br />In this paper we argue that it is possible and desirable to combine the easy input affordances of text with the powerful retrieval and visualization capabilities of graphical applications. We present WenSo, a tool that uses lightweight text input to capture richly structured information for later retrieval and navigation in a graphical environment..<br />Rambling E-mail<br />A previous board member, Steve Burleigh, created our web site last year and gave me alot of ideas. For this year, I found a web site called eTeamZ that hosts web sites for sports groups. Check out our new page: […]<br />Highly Technical Writing<br />Figure 3 shows the pseudocode that implements this design for Lookup. FAWN-DS extracts two fields from the 160-bit key: the i low order bits of the key (the index bits) and the next 15 low order bits (the key fragment).<br />
  35. 35. Blog – 83%<br />Print publishers are in a tizzy over Apple’s new iPad because they hope to finally be able to charge for their digital editions. But in order to get people to pay for their magazine and newspaper apps, they are going to have to offer something different that readers cannot get at the newsstand or on the open Web.<br />Classic uist– 87%<br />The metaDESK effort is part of the larger Tangible Bits project. The Tangible Bits vision paper, which introduced the metaDESKalong withand two companion platforms, the transBOARD and ambientROOM.<br />Draft uist– 90%<br />In this paper we argue that it is possible and desirable to combine the easy input affordances of text with the powerful retrieval and visualization capabilities of graphical applications. We present WenSo, a tool thatwhich uses lightweight text input to capture richly structured information for later retrieval and navigation in a graphical environment..<br />Rambling E-mail – 78%<br />A previous board member, Steve Burleigh, created our web site last year and gave me alot of ideas. For this year, I found a web site called eTeamZ that hosts web sites for sports groups. Check out our new page: […]<br />Technical Computer Science – 82%<br />Figure 3 shows the pseudocode that implements this design for Lookup. FAWN-DS extracts two fields from the 160-bit key: the i low order bits of the key(the index bits) and the next 15 low order bits (the key fragment).<br />
  36. 36. Average Performance<br />Cost: $1.41 per paragraph $0.55 to Find an average of two patches $0.48 to Fix each patch $0.38 to Verify the results<br />Time: Wait : median 18.5 minutes (Q1 = 8.3 min, Q3 = 41.6 min)<br /> Work: median 2.0 minutes<br /> (Q1 = 60 sec, Q3 = 3.6 min)<br />
  37. 37. Qualitative Observations<br />Works best with unnecessary text<br />[…] they are going to have to offer something different […]<br />Lack of domain knowledge[…] In this paper we argue that tangible interfaces […]<br />Parallel edits can be inconsistent<br />FAWN-DS extracts two fields from the 160-bit key: the i low order bits of the key (the index bits) and the next 15 low order bits (the key fragment).<br />

×