kwic: Extract Key Words in Context
In trinker/discon: Discourse Connectors Analysis

Description Usage Arguments Details Value References See Also Examples

Extract key words in context (kwic) utilizing freq_terms.

kwic(text.var, grouping.var, n.before = 1, tot = FALSE,
  n.after = n.before, ord.inds = TRUE, markup = c("<<", ">>"), top = 15,
  at.least = 5, extend = TRUE, ignore.case = FALSE, left = "\\b",
  right = "", names = c("KeyWords"), elim.old = FALSE,
  stopwords = c("going", qdapDictionaries::contractions[[1]],
  qdapDictionaries::Top200Words), ...)

`text.var`	The text variable.
`grouping.var`	The grouping variables. Also takes a single grouping variable or a list of 1 or more grouping variables.
`n.before`	The number of rows before the indexed occurrence.
`tot`	logical. If `TRUE` condenses sub-units (e.g., sentences) into turns of talk for that `grouping.var`.
`n.after`	The number of rows after the indexed occurrence.
`ord.inds`	logical. If `TRUE` inds is ordered least to greatest.
`markup`	A character vector of length two indicating the left (element 1) and right (element 2) boundary markers to use to highlight the key words. Use `c("", "")` to not mark the discourse connectors.
`top`	Top number of terms to show.
`at.least`	An integer indicating at least how many letters a word must be to be included in the output.
`extend`	logical. If `TRUE` the `top` argument is extended to any word that has the same frequency as the `top` word.
`ignore.case`	logical. If `TRUE` leading character's case is ignored.
`left`	The left boundary (regular expression) to put on the left side of words. Default is word boundary (i.e., `\\b`) that cuts off at non-word characters.
`right`	The right boundary (regular expression) to put on the right side of words. Default is no boundary and will locate words within words. For example the expression `"\\blast"` will find "last" and "lastly" but not "blast". To capture precisely key words use `right = "\\b"` or `right = left`.
`names`	A string indicating the name of the key words in the collective group name for the key words for use in the `termco` and generic `plot` (`dispersion_plot`) functions.
`elim.old`	logical. If `TRUE` eliminates the columns that are combined together by the named match.list.
`stopwords`	A character vector of words to remove from the text.
`...`	Other arguments passed to `termco`.

While kwic does not analyze discourse connectors, it is useful in identifying themes (and distribution across time) and builds upon the modular use of functions in the pkgdiscon package.

Returns returns a list of 2-3:

`counts`	A `termco` object of discourse connector counts.
`Context 1`	A `trans_context` object of the key words in context. Note the name of this object is supplied by `names` element one.
`Context 2...n`	An optional (not returned if `regex` is of length one) `trans_context` object of the key words in context. Note the name of this (these) object(s) is supplied by `names` element 2...n.

Ryan, G. W. & Bernard, H. R. (2003). Techniques to identify themes. Field Methods. 15(1), 85-109. doi: 10.1177/1525822X02239569

termco, trans_context, freq_terms

out <- with(pres_debates2012, kwic(dialogue, list(time, person)))
plot(out[[1]])
head(out[[2]])
with(pres_debates2012, plot(out, grouping.var = person, rm.vars = time,
    total.color = NULL))

## Save externally use .doc or .txt
## print(out[[2]], file="kwic.doc")