tworavens: a graphical, browser-based statistical interface for data repositories by vito d’orazio...
TRANSCRIPT
![Page 1: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/1.jpg)
TwoRavensA Graphical, Browser-Based Statistical Interface
for Data Repositories
Vito D’Orazio and James Honaker
Data ScienceInstitute for Quantitative Social Science
Harvard University
June 11, 2015
![Page 2: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/2.jpg)
![Page 3: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/3.jpg)
TwoRavens• Gesture-based web application• Explore and analyze tabular data files
I Nearly 25,000 on Harvard’s Dataverse
• Easy access to descriptive statistics• Interactive statistical modeling using the language of
directed graphs
![Page 4: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/4.jpg)
TwoRavens• Gesture-based web application• Explore and analyze tabular data files
I Nearly 25,000 on Harvard’s Dataverse• Easy access to descriptive statistics• Interactive statistical modeling using the language of
directed graphs
![Page 5: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/5.jpg)
TwoRavens• Gesture-based web application• Explore and analyze tabular data files
I Nearly 25,000 on Harvard’s Dataverse• Easy access to descriptive statistics• Interactive statistical modeling using the language of
directed graphs
![Page 6: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/6.jpg)
TwoRavens• Gesture-based web application• Explore and analyze tabular data files
I Nearly 25,000 on Harvard’s Dataverse• Easy access to descriptive statistics• Interactive statistical modeling using the language of
directed graphs
![Page 7: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/7.jpg)
Design Principles
• Browser-based, thin clientI Data are never localI Metadata are pulled client-side
I Dataverse’s DDI metadataI TwoRavens’ generated metadata
I Data are pulled server-side• Device independent and broadly accessible
I Only requires an internet connectionI Does not presuppose expert statistical knowledge or
experience with statistical software
![Page 8: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/8.jpg)
Architecture
![Page 10: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/10.jpg)
Future Directions
1. Automated Statistical Model Selection2. Accumulation of User Results3. Interface for Curator Privacy Model
![Page 11: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/11.jpg)
Future Directions
1. Automated Statistical Model Selection
2. Accumulation of User Results3. Interface for Curator Privacy Model
![Page 12: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/12.jpg)
Future Directions
1. Automated Statistical Model Selection2. Accumulation of User Results
3. Interface for Curator Privacy Model
![Page 13: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/13.jpg)
Future Directions
1. Automated Statistical Model Selection2. Accumulation of User Results3. Interface for Curator Privacy Model
![Page 14: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/14.jpg)
PrivateData Curator
PUser
PUser
PUser
q1
p1
q2
p2
p3
q3
Figure: The curator architecture for data privacy.
![Page 15: TwoRavens: A Graphical, Browser-Based Statistical Interface for Data Repositories by Vito D’Orazio and James Honaker](https://reader033.vdocument.in/reader033/viewer/2022042819/55cef21abb61ebc53d8b45a2/html5/thumbnails/15.jpg)