Note: This metadata was prepared by the CRAWDAD team and verified by the data set (or tool) authors. We have made every effort to ensure its accuracy, but urge all users to consider the metadata and data carefully and be sure that their use in research is consistent with the nature and limitations of the data. We welcome any corrections.
This metadata was prepared based on the following reference(s):
|
version
| v. 2011-03-23 |
|
changes
| the initial version |
|
bibtex
|
@MISC{strath-nodobo-2011-03-23,
author = {Alisdair McDiarmid and James Irvine and Stephen Bell and Jamie Banford},
title = {{CRAWDAD} data set strath/nodobo (v. 2011-03-23)},
howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/strath/nodobo},
month = mar,
year = 2011
}
|
| metadata last modified | 2011-07-05 |
| summary | Dataset gathered by Nodobo, a suite of social sensor software for Android
phones, during a study of the mobile phone usage at University of Strathclyde. |
| release date | 2011-03-23 |
| measurement start | 2010-09-09 |
| measurement end | 2011-02-23 |
| authors | Alisdair McDiarmid
James Irvine
Stephen Bell
Jamie Banford
|
| license | Copyright (c) 2011 University of Strathclyde
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE. |
|
web site
| http://www.crawdad.org/strath/nodobo |
|
wiki
|
go to the wiki page for this data set
|
| keyword | Bluetooth, cellular network, Wi-Fi hotspot, location, social network |
| measurement purposes | Usage Characterization
Social Network Analysis
|
| network type | bluetooth |
| network type | GSM (Global System for Mobile Communications) |
| network type | cellular network |
| environment | Our researchers developed "Nodobo", a set of software extensions to the Google
Android operating system, for enabling the capture and replay of smartphone
user interactions sessions.
The software captures a variety of social context data, including logs of phone
calls, text messages, Bluetooth proximity detection, WiFi access point, and
cell tower ID. The directionality of calls and text messages are recorded,
along with the associated phone number, and the duration of the call or length
of the message. Bluetooth proximity is detected every minute, and includes all
devices in the study as well as any other clients which respond to service
discovery. Basic positioning is achieved through WiFi hotspot and cell tower
ID records. |
| network | Each of the study participants was given a Google Nexus One smartphone,
prepared with a modified Android operating system. Data is stored in a simple
database on the device SD card, which is then synchronised over the air
to a central server. |
| collection | The dataset was collected through monitoring devices of 27 users over a
5-month study. |
| sanitization | Record fields containing personally identifiable information have been
anonymised. |
|
tracesets included
| strath/nodobo/mobile (v. 2011-03-23)
|
|
version
| v. 2011-03-23 |
|
changes
| the initial version. |
|
bibtex
|
@MISC{strath-nodobo-mobile-2011-03-23,
author = {Alisdair McDiarmid and James Irvine and Stephen Bell and Jamie Banford},
title = {{CRAWDAD} trace set strath/nodobo/mobile (v. 2011-03-23)},
howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/strath/nodobo/mobile},
month = mar,
year = 2011
}
|
| metadata last modified | 2011-07-05 |
| summary | Nodobo-2011-01-v1 is the traceset gathered by Nodobo software at University
of Strathclyde from September 2010 to February 2011. |
| release date | 2011-03-23 |
| measurement start | 2010-09-09 |
| measurement end | 2011-02-23 |
| measurement purposes | Usage Characterization
Social Network Analysis
|
| methodology | A group of 27 promising high school students in a Scottish state high school
were selected for this study. All students previously had a mobile phone, with
approximately 1/3 of these falling in the category of smartphone (iPhone,
Blackberry, or similarly powerful handset). Each of the study participants was
given a Google Nexus One smartphone, prepared with a modified Android
operating system.
The close proximity of the deployment to University of Strathclyde enables the
study organisers to schedule regular visits to diagnose issues, as well as
facilitating regular backups to be made. To maintain as up-to-date a dataset
as possible, and to limit the number of visits required, the devices also
synchronise with a web server over the mobile network or WiFi. |
| sanitization | Record fields containing personally identifiable information have been
anonymised. |
| download url | Download (41MB gz)
(MD5 Hash: 018004f5f9bab186c2815058e6d732d1) from US UK AU |
| download url | Download (32MB gz)
(MD5 Hash: 1ff9a236b96f9c3dfe7873d01e9b2204) from US UK AU |
| parent data | strath/nodobo (v. 2011-03-23)
|
|
traces included
| strath/nodobo/mobile/social (v. 2011-03-23)
|
|
version
| v. 2011-03-23 |
|
changes
| the initial version |
|
bibtex
|
@MISC{strath-nodobo-mobile-social-2011-03-23,
author = {Alisdair McDiarmid and James Irvine and Stephen Bell and Jamie Banford},
title = {{CRAWDAD} trace strath/nodobo/mobile/social (v. 2011-03-23)},
howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/strath/nodobo/mobile/social},
month = mar,
year = 2011
}
|
| metadata last modified | 2011-07-05 |
| summary | Data gathered by Nodobo software during a study of the mobile phone usage
at University of Strathclyde from September 2010 to February 2011. |
| derived | false |
| release date | 2011-03-23 |
| measurement start | 2010-09-09 |
| measurement end | 2011-02-23 |
| configuration | 27 Google Nexus One smartphones were prepared with a modified Android
operating system, running Nodobo. The phone database is synchronised
periodically over-the-air with a web services data store. |
| format | db.sqlite3.dump.bz2 is a bzipped SQL dump of the sqlite3 database. You can
recreate the database by doing the following:
bzcat db.sqlite3.dump.bz2 | sqlite3 db.sqlite3
# Database schema
The following tables are used:
## Calls and Messages
* other_id: id of the other user on the call (NULL if not in the study)
* number: phone number of the other end of the call/message (related:
Users#number)
* duration: length of the call in seconds
* length: number of characters in the message
## CellTowers
* cellid: GSM base transceiver station CID
* lac: location area code
## Devices
* imei: blank for this release of the data
* mac: Bluetooth MAC (related: Presences#mac)
## Presences
* other_id: user_id of the detected device (NULL if not in the study)
* mac: Bluetooth MAC (related: Devices#mac)
* bluetooth_class: reported class of the device
* name: human-readable name of the device
## Users
* name: "Anonymous" for this release of the data
* number: phone number of the study user (related: Calls#number,
Messages#number)
## Wifis
* ssid: human-readable name of the base station
* bssid: base station MAC
## All tables
* The database schema follows ActiveRecord conventions: tables are plurals,
foreign keys are singular_id, each table has an id primary key and
created_at/updated_at timestamps.
* user_id is used to indicate which user recorded the interaction.
* Calls and messages tables have two timestamp columns. The
call_timestamp/message_timestamp is the one recorded by the phone when the
call/message was originally recorded. The timestamp column in the time at
which the calldb/smsdb synchronisation occurred (which is less useful).
* Some tables have an "interaction" column. This was used for database
synchronising and is left in for internal debugging purposes.
# Software and studies
Also included in the dataset download are programs for three sample studies. These are detailed below.
Each program can be run with ruby: for example, "ruby conversation-length.rb". The programs assume that your current working directory is the one with the database and the nodobo.rb code.
Software used:
* Ruby 1.8.7 or later, with gems: activerecord, sqlite3-ruby, progressbar
* gnuplot 4.4
* GraphViz 2.22
## Ruby interface: nodobo.rb
We have supplied a simple ActiveRecord interface to the database, "nodobo.rb". This gives classes and relations for each of the types of data in the dataset.
The interface can be used by running "irb -r ./nodobo.rb", or by using "require 'nodobo'" in your own programs. A sample irb session is given below:
>> u = User.find(19)
=> #<User id: 19, name: "Anonymous", number: "07102745960", created_at: "2010-11-11 10:19:34", updated_at: "2010-11-11 10:19:34">
>> u.calls.size
=> 976
>> study_calls = u.calls.select {|c| c.other != nil }; study_calls.size
=> 133
>> Hash[study_calls.group_by(&:other_id).map {|k,v| [k, v.size]}]
=> {16=>2, 19=>1, 25=>2, 14=>4, 21=>124}
>> v = User.find(21)
=> #<User id: 21, name: "Anonymous", number: "07456622368", created_at: "2010-11-11 10:19:35", updated_at: "2010-11-11 10:19:35">
>> v.calls.select {|c| c.other != nil }.size
=> 175 |
| sanitization | The following fields have been altered to remove personal information from the
dataset:
* Call#number, Message#number, User#number
* Device#mac, Presence#mac
* Wifi#bssid
* Presence#name
* Wifi#ssid
* CellTower#cellid
* CellTower#lac
Each real value for these fields maps 1:1 to a randomly-generated anonymous
value. The process for generating these values is as follows:
* Phone number: random number with the same number of digits; if original
number is 3 or more digits, keep the original first 2 digits
* MAC address: 12 random hex digits
* Bluetooth name/Wifi ssid: random sequence of dictionary words, same number
of words as original name
* Cell ID and LAC: random number with the same number of digits |
| parent data | strath/nodobo/mobile (v. 2011-03-23)
|
