(094) previous ~ index ~ next

---

To: "Jonathan Fiscus" <jonathan.fiscus@nist.gov>,
From: "George Doddington" <doddington@email.msn.com>
Subject: Re: Type of TDT2 manual transcription data => CC
Date: Wed, 15 Jul 1998 19:46:14 -0400

---

I would like to reiterate what Fiscus has said, namely that the type of
manual transcription data will be closed caption data. The primary reason
for choosing CC data is that it is has the best coverage of the TDT2 corpus.
We didn't want the manual transcription to be a hybrid of different types of
data, so we decided to use CC data. The FDCH data will be used only as a
special ancillary test that lies outside of the conditions published in
the TDT2 evaluation plan.
-----------------------
George R. Doddington -- reply to doddington@sri.com, or call 703/556-3434

---

(094) previous ~ index ~ next

Last updated Wed Sep 9 09:40:52 1998