2 thoughts on “Extracting speaker meta-data from the BNC

    tiflo said:
    July 1, 2010 at 6:57 pm

    Hey Judith,

    could you post some example sentence to give people an idea of the format of the files? Thanks to you and Ben for putting this together.

    Florian

    Like

    vandurme said:
    July 2, 2010 at 7:46 am

    A few more tidbits:

    It was parsed with the default settings of the Charniak-Johnson parser,
    in the winter of ’08-09, using the University of Rochester BlueHive
    cluster.

    Total # of sentences in corpus: roughly 6 million (6042908)
    Total # of sentences in written: roughly 5 million (5007381)
    Total # of sentences in spoken: roughly 1 million (1035527)

    Total # of tokens in corpus: roughly 112 million (116940003)
    Total # of tokens in written: roughly 100 million (104603588)
    Total # of tokens in spoken: roughly 12 million (12336415)

    Like

Questions? Thoughts?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s