Internet General Inquirer
Internet General Inquirer
Welcome to the Internet Java version of the General Inquirer, which offers
the content analysis resources originally described in these two books:
- Philip J. Stone, Dexter C. Dunphy, Marshall S. Smith, Daniel M. Ogilvie,
and associates. "The General Inquirer: A Computer Approach to Content
Analysis," The MIT Press, 1966.
- Edward Kelly and Philip Stone. "Computer Recognition of English
Word Senses." North-Holland Linguistic Series, 1975.
This version uses the H4-4 tag categories developed by Dexter Dunphy
and colleagues at the University of New South Wales, with additional
categories and entries recently created by Stacey Di Cicco at Harvard.
The current system identifies about 13,000 word roots and utilizes
6,336 disaambiguation rules. This Java version was written and is
maintained by Vanja Buvac
(buvac@fas.harvard.edu).
Current support is provided by The
Gallup Organization.
Instructions
To test out the General Inquirer, please type or paste in one or more text
sentences that you want content analyzed in the box area below. Choose the
level of detail you want the program to display by marking the appropriate
checkbox. The "Show disambiguation trace" option will display a trace of
the disambiguating process. This option produces a large output file, and
should be used only on relatively short texts. Checking the "Show sentence
summaries" option will display the tags assigned to each word in each
sentence, as well as the sense number assigned to each disambiguated word
root. Note that the Inquirer does not first attempt to correct spelling.
Click "submit" to process all the text in the box.
Cumulative statistics for the whole text are displayed at the end of
the output. Each content analysis tag assigned at least once in the text
is listed, together with its raw count and the percent of words in the text
having that tag. In addition, all the words or word senses in the text that
were assigned that tag are listed, together with their frequencies if
greater than one. In newer versions of Netscape and Internet Explorer (v.5
on pc, v. 4.5 on Mac), you can simply move the cursor over a word to get
the definition of the word displayed in a small "balloon." If the word is
in bold font, you will obtain a definition of the word, including the
frequency percent of a word sense. If the word is not in bold font, you
will get all the tags assigned to that word. Moving a cursor over an
abbreviated tag name will give the full tag name, while mousing that tag
name will open a window showing all the words or word senses in the system
that are assigned that tag.
Click here
to get an alphabetical list of all the tags in the
dictionary. Click here to get a list
of tags and the associated words that the current program is tracking.
Try it out: