Changes between Version 5 and Version 6 of BossaReference
- Timestamp:
- Jul 18, 2008, 9:17:15 PM (17 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
BossaReference
v5 v6 34 34 These structures are stored in the database. 35 35 36 == Volunteer characteristics == 36 For jobs, the app data typically contains the information needed to display 37 the job to a user - e.g., the name of an image file. 38 For calibration jobs, it would also contain the correct answer. 37 39 38 For each application and each user, 39 Bossa maintains ''skill estimate'', an estimate of the user's skill at that task. 40 This is maintained in the user's database record. 41 Normally it's a single number in [0..1], and it's initially zero. 40 For instances, the app data typically contains the user's response to the job. 42 41 43 The skill estimate can be computed in any of several ways: 42 For users, the app data typically contains a representation of the 43 user's "skill", possibly including: 44 44 45 * The results of the user's interaction with a Bolt course associated with the application.46 * The user's performance on "calibration jobs" mixed into the stream.47 * The fraction of the user's results classified as invalid by redundancy.45 * their score on a training course 46 * their performance on calibration tasks 47 * their performance based on validation 48 48 49 Skill estimates are used for two purposes: 49 If a project has multiple applications, it should keep separate 50 data for each applications. 50 51 51 * To decide whether to give jobs to a user; 52 * To decide how many redundant instances of a given job are needed. 53 52 == User categories == 54 53 55 54