Web Scale Collaborations

Transcript

1. Web Scale Collaborations Arfon Smith @arfon
2. Citizen Science
3. Distributed Computing
5. Distributed Data Collection
8. Distributed Analysis
13. http://www.novacelestia.com
19. 0 250,000 500,000 750,000 1,000,000 Professor Paper PhD SDSS
20. Classifications per hour 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 Hours 0 6 12 18 24 30 36 42 48 1 Kevin months Fukugita et al. 2007
30. SDSS HST Starforming pea Narrow-line Seyfert pea
38. Motivations
42. 1,000,000,000,000 hours / year
43. Spectrum of cognitive surplus
46. Begins with open data
47. Open Source
48. Not treating code and data as first class research objects
49. GitHub
50. What is a GitHub?
55. Easier to work together than alone
56. Open Source collaboration
73. Open Public ≠
74. Open (within your team, department or institution)
75. Electronic
76. Available
77. Asynchronous
78. Lock-free
82. Low friction collaboration
83. “open source is… reproducible by necessity” Fernando Perez http://blog.fperez.org/2013/11/an-ambitious-experiment-in-data-science.html
84. Better at collaborating because they have to be
85. Towards Collaborative Versioned Science
86. How do we make this behaviour the norm?
87. Incentive model (it’s broken)
88. Credit
89. http://dx.doi.org/10.6084/m9.figshare.828487
90. http://dx.doi.org/10.6084/m9.figshare.828487
93. Derive meaningful metrics from open contributions
94. “Academic environments of today do not reward tool builders” Ed Lazowska, OSTP event http://lazowska.cs.washington.edu/MS/MS.OSTP.pdf
95. A VISION AND STRATEGY FOR SOFTWARE FOR SCIENCE, ENGINEERING, AND EDUCATION
96. What can we do today?
97. Take data management plans seriously
98. Try versioning your research
99. Share more than just data
100. If you’re going to share it then you better put a licence on it
101. Thanks. arfon@github.com @arfon "