While the context message places the actions of the student in
context of the curriculum, the tool message captures the actions of the
student. Generate a <tool_message>
whenever the student interacts with the media or the tutor itself
performs some action on the media. This will capture whenever media is
stopped, started, muted, etc.
A tool message in media logging is structurally the same as a tool message for tutor logging. See Section 2.5, “<tool_message>” for a description of the tool message content model. In addition, the following constraints exist.
Each tool message must include a <semantic_event> element. This element
contains the transaction_id, which must
be unique within the context message for each tool message. If this is
a tutored event, a subsequent <tutor_message> will be paired with the
tool message by this transaction_id.
![[Note]](images/note.gif) | Note |
|---|---|
|
We decided that the media event is semantically important and
as such we use the |
The name attribute of the <semantic_event> is dependent on whether
or not the event is tutored. A tutored event uses the standard ATTEMPT value; an untutored event should use a
value from the following table:
Table 5. Recommended values for the <semantic_event> name attribute for a media object
name
attribute value | Description |
|---|---|
VIDEO_ACTION | Used for an action on a video object. |
AUDIO_ACTION | Used for an action on an audio object. |
MEDIA_ACTION | Used for an action on another type of media object. |
The <semantic_event> element
has two attributes which are worth noting here: the trigger and the subtype. The trigger should be set to either
of the following:
Table 6. Recommended values forthea <semantic_event> trigger attribute for a media object
trigger
attribute value | Description |
|---|---|
USER | If the user initiates the event. |
DATA | If the tool or tutor initiates the event. |
The subtype attribute is more of a
free-form field that further qualifies the type of event. It can be
set to something like tutor-performed. The
value is stored in the DataShop database and exported, but no
reasoning is dependent on this value.
Included in each tool message is the <event_descriptor> element which provides
the details of the semantic event: it describes the specifics of the
student's interaction with the tool. This should be a measurable
change of state in the tool. It can be either a student action or an
action taken for the student by the tool (see Section 3.3.1, “<semantic_event> (media logging)” for how to signify
events initiated by the tutor).
For media logging, the event descriptor will include the action
performed, the time(s) it occurred within the clip, and which
interface elements were part of the interaction. Place these in the
<selection>, <action>, <input> elements, each with a type attribute to further clarify the
information those elements contain. The recommended list of media
actions along with their expected selections and inputs are as
follows:
Table 7. Format of an <event_descriptor> selections, action, and input values for a media object
| Selections | Action | Input(s) | Notes |
|---|---|---|---|
| play | time | start/resume playback of clip; time is location in clip where playback was started |
| pause | time | time is location in clip where playback was paused |
| stop | time | time is location in clip where playback was stopped |
| close | time | time is location in current clip when media was closed |
| end | time | the end of the clip has been reached and playback has stopped; time should be the same as clip length |
| cue |
| start_cue is the time in the clip when the starting cue point was added; end_cue is the time in the clip when the ending cue point was added |
| mute | time | time is location in clip when this action occurred |
| unmute | time | time is location in clip when this action occurred |
| volume_change |
| time is location in clip when this action occurred. start_level is the volume level before the change, and final_level is the resulting volume level. |
| next_clip | time | move to the next media clip; time is location in current clip when this action occurred |
| prev_clip | time | move to the previous media clip; time is location in current clip when this action occurred |
Every action is expected to have three selections. The first selection is the widget, button, or component that the student actually clicked on. The second is the name of the media file, and the third is the length of that file.
Example 3. Three selections for every media action
<selection>_level0.VideoPlayerInstance1.sliderButtonName</selection> <selection type="media_file">mymovie.flv</selection> <selection type="clip_length">00:08:34.823</selection>
The <input> elements will be
used to indicate the media timing. Similar to <selection> elements the type attribute will be used to differentiate
the different inputs. The inputs will vary by the action as noted in
the table above. All time values should be in the form
HH:MM:SS.SSS
The following are examples of typical media event log messages.
Example 5. Tool message for "play"
<tool_message
context_message_id ="02CE3AE5-F6D5-9177-913F-C34730F1096C">
<semantic_event
transaction_id="1F3A9B23-9164-DD83-EBB2-1589FD38D4B3"
name="VIDEO_ACTION" />
<event_descriptor>
<selection>_level0.VideoPlayerInstance1.sliderButtonName</selection>
<selection type="media_file">mymovie.flv</selection>
<selection type="clip_length">00:08:00.0</input>
<action>play</action>
<input type="time">00:02:34.2</input>
</event_descriptor>
</tool_message>
Example 6. Tool message for "stop"
<tool_message
context_message_id ="02CE3AE5-F6D5-9177-913F-C34730F1096C">
<semantic_event
transaction_id="A43B003-9164-DD83-EBB2-1589FD38D435"
name="VIDEO_ACTION" />
<event_descriptor>
<selection>_level0.VideoPlayerInstance1.sliderButtonName</selection>
<selection type="media_file">mymovie.flv</selection>
<selection type="clip_length">00:08:00.0</input>
<action>stop</action>
<input type="time">00:05:32.2</input>
</event_descriptor>
</tool_message>
Example 7. Tool message for "cue"
<tool_message
context_message_id ="02CE3AE5-F6D5-9177-913F-C34730F1096C">
<semantic_event
transaction_id="B503948-9164-DD83-EBB2-1589FD38D435"
name="VIDEO_ACTION" />
<event_descriptor>
<selection>_level0.VideoPlayerInstance1.sliderButtonName</selection>
<selection type="media_file">mymovie.flv</selection>
<selection type="clip_length">00:08:00.0</input>
<action>cue</action>
<input type="start_cue">00:04:34.8</input>
<input type="stop_cue">00:05:42.2</input>
</event_descriptor>
</tool_message>