UNIX_user_data
OpenML dataset with id 373
No author found.
Full work available at URL: https://api.openml.org/data/v1/download/52413/UNIX_user_data.arff
Upload date: 27 September 2014
Dataset Characteristics
Number of classes: 9
Number of features: 3 (numeric: 1, symbolic: 1 and in total binary: 0 )
Number of instances: 9,100
Number of instances with missing values: 0
Number of missing values: 0
Author: Terran Lane (terran@ecn.purdue.edu) Source: UCI - Date unknown Please cite:
This file contains 9 sets of sanitized user data drawn from the command histories of 8 UNIX computer users at Purdue over the course of up to 2 years (USER0 and USER1 were generated by the same person, working on different platforms and different projects). The data is drawn from tcsh(1) history files and has been parsed and sanitized to remove filenames, user names, directory structures, web addresses, host names, and other possibly identifying items. Command names, flags, and shell metacharacters have been preserved. Additionally, SOF and EOF tokens have been inserted at the start and end of shell sessions, respectively. Sessions are concatenated by date order and tokens appear in the order issued within the shell session, but no timestamps are included in this data. For example, the two sessions:
cd ~/private/docs ls -laF | more cat foo.txt bar.txt zorch.txt > somewhere exit
cd ~/games/ xquake & fg vi scores.txt mailx john_doe@somewhere.com exit
would be represented by the token stream
SOF cd \<1\> (one "file name" argument) ls -laF | more cat \<3\> (three "file" arguments) \> \<1\> exit EOF SOF cd \<1\> xquake & fg vi \<1\> mailx \<1\> exit EOF
This data is made available under conditions of anonymity for the contributing users and may be used for research purposes only. Summaries and research results employing this data may be published, but literal tokens or token sequences from the data may not be published except with express consent of the originators of the data. No portion of this data may be released with or included in a commercial product, nor may any portion of this data be sold or redistributed for profit or as part of of a profit-making endeavor.
This page was built for dataset: UNIX_user_data
