Mark (.mrk) file specifications, Updated 03/29/94:

The mark files contain time-aligned word transcriptions.

1. In .mrk files, each record has 4 fields; the first indicates the
talker (A or B), followed by a turn indicator, A.x or B.x (where x
indicates the speaker turn); the second field contains the estimated
start time of the current word, the third field contains the duration
of the word, and the fourth contains the the word.

2. The first field may contain the dummy symbol "*" when the event in
field 4 is not attributed to either speaker.  Examples: "[beep]" at
the beginning indicating the dtmf tone, or the "..." at the end
indicating that the conversation was cut off (timed out).

  *        *       *    [Beep]
 @A.5     1.36    0.28  Okay,
 @A.5     1.64    0.08  I


3. Fields 2 and 3 may contain the dummy symbol "*" when the event in
field 4 is a nonspeech event, a comment, or a stand-alone string of
punctuation (i.e., something which does not receive a duration from
the time alignment algorithm), as well as the cases in (2) above.

  A.3        *       *  [lipsmack]
  A.3        *       *  {pause}
  A.3   311.02    0.62  economic


4. Fields 2 and 3 may contain the dummy symbol "*" when there is
simultaneous speech by A and B.  In this case the time alignment algorithm was
allowed to recognize the speech of EITHER talker--usually the louder
one.  The other talker's speech is not time aligned during this
period, but the words should be attributed to A and B correctly.

  A.7   113.96    0.24  thing
  A.7   114.20    0.10  is
  A.7   114.30    0.44  still,
  A.7        *       *  {pause}
  A.7        *       *  #you
  A.7        *       *  know#
  A.7        *       *  --
  B.8   116.40    0.20  #Your
  B.8   116.60    0.56  education.#
  A.9        *       *  --
  A.9   117.16    0.22  getting
  A.9   117.38    0.10  your
  A.9   117.48    0.60  education.


5. In field 1, "@" or "@@" may occur before A or B to mark suggested
trim points for topicality purposes.  Although "@" may be present in
some (*.mrk) files, they are irrelevent for the purpose of speaker
identification and may be ignored.