Difference between revisions of "DiHabs"

From OpenScience
Jump to: navigation, search
m (Datasets)
(Datasets)
Line 12: Line 12:
 
!Survey
 
!Survey
 
!Description
 
!Description
 +
|Date
 
|-
 
|-
| TODO
+
| [http://wwwiuk.informatik.uni-rostock.de/~cb203/dihabs-institute-survey.db Institute]
| TODO
+
| Results of a small examplary test survey (10 participants).
 +
| October 2010
 +
|-
 +
| [http://wwwiuk.informatik.uni-rostock.de/~cb203/dihabs-test-survey.db Test]
 +
| Results of a survey with around 70 participants at the computer science institute at Rostock university.
 +
| September 2010
 
<!--
 
<!--
 
|[http://wwwiuk.informatik.uni-rostock.de/~cb203/dihabs-example-survey.rdb Example survey]
 
|[http://wwwiuk.informatik.uni-rostock.de/~cb203/dihabs-example-survey.rdb Example survey]
Line 23: Line 29:
 
==== Format ====
 
==== Format ====
  
''DiHabs'' survey results are stored as [http://docs.python.org/library/shelve.html Python shelve] databases. The analyzer script within the ''DiHabs'' application (see [[DiHabs#Software|Software]] below) dumps its content in limited form. For full access please refer to the linked Python documentation.
+
''DiHabs'' survey results are stored as [http://docs.python.org/library/shelve.html Python shelve] databases. A result database is structured as a dictionary where each item represents one survey participant. The analyzer script within the ''DiHabs'' application (see [[DiHabs#Software|Software]] below) dumps its content in limited form and generates various statistical plots. Details about what information is stored per participant can be found in the file <code>analyze.py</code> within the ''DiHabs'' application.
 
+
A result database is structured as a dictionary where each items represents one survey participant. Per participant the following information is stored:
+
 
+
{| cellpadding=5 border=1
+
!interview
+
|Complete interview specification.
+
|-
+
!end
+
|Time in seconds the participant needed to answer all questions.
+
|-
+
!npersons
+
|Number of persons entered by a participant in the first interview part.
+
|-
+
!groups
+
|Generated person groups for the third interview part.
+
|-
+
!nvalues
+
|List of number of items entered by participants entered in the second interview part for each information type specified in the interview.
+
|-
+
!disclosures
+
|List of answers from the third interview part. Each answer is also a dictionary containing items which specify the corresponding person group, disclosed values at all and per type and the progressed time since the interview started.
+
|}
+
  
 
=== Publications ===
 
=== Publications ===

Revision as of 09:44, 20 October 2010

Project Description

DiHabs is a survey system to capture and analyze information disclosure habits. It has been developed within the project Privacy in Smart Environments.

Contact contributers

Datasets

Survey Description Date
Institute Results of a small examplary test survey (10 participants). October 2010
Test Results of a survey with around 70 participants at the computer science institute at Rostock university. September 2010

Format

DiHabs survey results are stored as Python shelve databases. A result database is structured as a dictionary where each item represents one survey participant. The analyzer script within the DiHabs application (see Software below) dumps its content in limited form and generates various statistical plots. Details about what information is stored per participant can be found in the file analyze.py within the DiHabs application.

Publications

A paper presenting DiHabs has been submitted to EuroSys 2011.

Software

DiHabs is implemented as a Django web application. The download files below contain this web application integrated in a development Django site project (for easy testing outside productive server environments). Documentation is not yet complete. Please contact me directly for installation and usage support.

Version Date Description
TODO TODO TODO