| 24 | | == Abstractions == |
| 25 | | |
| 26 | | A Bossa project has one or more ''skill apps''. |
| 27 | | A skill app has a dynamic set of ''skill tasks''. |
| 28 | | Each one has an associated set of arguments describing its parameters or input files. |
| 29 | | Each skill task has a set of ''task instances''. |
| 30 | | Each one represents a copy of the task, either in progress or completed. |
| 31 | | Each instance is assigned either to a user or to a team. |
| 32 | | |
| 33 | | Skill apps are classified as: |
| 34 | | |
| 35 | | * Online: the task is performed by a single user, sequentially, in a web browser. |
| 36 | | * Offline: the task is not online, e.g. because it's potentially handled by a group of users, or requires other asynchronous activity. |
| 37 | | |
| 38 | | An app has an associated URL |
| 39 | | identifying a script that takes an task ID argument and displays the task instance. |
| 40 | | The task may consist either of a single web page or a sequence of web pages. |
| 41 | | In either case the last page in the sequence, when done, should call Bossa API |
| 42 | | functions to record the completion of the task, and perhaps display another task. |
| 43 | | |
| 44 | | Skill apps are either: |
| 45 | | |
| 46 | | * Individually validated: the app has a server-side program that examines a completed instance and decides if it's valid. |
| 47 | | * Group validated: the app has a server-side program that examines a group of instances, sees if there's a consensus, and if so constructs a ''canonical result'' and marks the instances as valid or invalid. |
| 48 | | |
| 49 | | A project can configure: |
| 50 | | |
| 51 | | * A maximum number of outstanding offline tasks per user or group |
| 52 | | * A maximum number of tasks per day issued per user or group |
| 53 | | |
| 54 | | == Volunteer characteristics == |
| 55 | | |
| 56 | | For each skill app and each user, |
| 57 | | Bossa maintains ''skill estimate'', an estimate of the user's skill at that task. |
| 58 | | This is maintained in the user's project-specific XML document. |
| 59 | | Normally it's a single number in [0..1], and it's initially zero. |
| 60 | | |
| 61 | | The skill estimated can be computed in any of several ways: |
| 62 | | |
| 63 | | * The results of the user's interaction with a Bolt course associated with the skill app. |
| 64 | | * The user's performance on "calibration tasks" mixed into the stream. |
| 65 | | * The fraction of the user's results classified as invalid by redundancy. |
| 66 | | |
| 67 | | Skill estimates are used for two purposes: |
| 68 | | |
| 69 | | * To decide whether to give tasks to a user; |
| 70 | | * To decide how many redundant instances of a given task are needed. |
| 71 | | |
| 72 | | == Implementation == |
| 73 | | |
| 74 | | To get work, a user goes to a particular Bossa-supplied page. |
| 75 | | There he sees a list of skill apps for which tasks are available |
| 76 | | and for which he is qualified, |
| 77 | | and links to courses for other apps. |
| 78 | | Online and offline apps are listed separately. |
| 79 | | Each app has an estimate of the time or other resources required to complete the task. |
| 80 | | |
| 81 | | Selecting an online app invokes the Bossa ''scheduler'' script, |
| 82 | | which selects a task instance suitable for the user, |
| 83 | | and redirects to its instance URL. |
| 84 | | |
| 85 | | Selecting an offline app invokes the Bossa scheduler, |
| 86 | | which selects a task and redirects to its instance-start URL. |
| 87 | | |
| 88 | | Team administrators are provided with an interface for getting |
| 89 | | offline tasks for the team. |
| 90 | | The scheduler allows a team to get instances only for apps |
| 91 | | for which some team member has the required skill. |
| 92 | | |
| 93 | | Users and teams are provided with an interface for seeing |
| 94 | | a list of pending offline tasks. |
| 95 | | They can indicate that one of them is completed; |
| 96 | | this takes them to the instance-complete URL for that task. |
| 97 | | |
| 98 | | == Integration with BOINC == |
| 99 | | |
| 100 | | Some offline tasks may involve computation done through BOINC; |
| 101 | | i.e. if the task is assigned to a team, the computation is queued |
| 102 | | in the project's BOINC server and dispatched to members of the team. |
| 103 | | (Or if the task is assigned to a user with many computers, |
| 104 | | those computers are used). |
