weblmGenerateNextWords function

Returns the words most likely to follow a sequence of words.

This function returns the list of words (completions) most likely to follow a given sequence of words. The input string must be in ASCII format.

Internally, this function invokes the Microsoft Cognitive Services Web Language Model REST API documented at https://www.microsoft.com/cognitive-services/en-us/web-language-model-api/documentation.

You MUST have a valid Microsoft Cognitive Services account and an API key for this function to work properly. See https://www.microsoft.com/cognitive-services/en-us/pricing

for details.

weblmGenerateNextWords(precedingWords, modelToUse = "body",
  orderOfNgram = 5L, maxNumOfCandidatesReturned = 5L)

Arguments

• precedingWords: (character) Character string to retrieve completions for. Must be in ASCII format.
• modelToUse: (character) Which language model to use, supported values: "title", "anchor", "query", or "body" (optional, default: "body")
• orderOfNgram: (integer) Which order of N-gram to use, supported values: 1L, 2L, 3L, 4L, or 5L (optional, default: 5L)
• maxNumOfCandidatesReturned: (integer) Maximum number of candidates to return (optional, default: 5L)

Returns

An S3 object of the class weblm. The results are stored in the results dataframe inside this object. The dataframe contains the candidate words and their log(probability).

Examples

## Not run:

 tryCatch({

   # Generate next words
   wordCandidates <- weblmGenerateNextWords(
     precedingWords = "how are you",  # ASCII only
     modelToUse = "title",            # "title"|"anchor"|"query"(default)|"body"
     orderOfNgram = 4L,               # 1L|2L|3L|4L|5L(default)
     maxNumOfCandidatesReturned = 5L  # Default: 5L
   )

   # Class and structure of wordCandidates
   class(wordCandidates)
   #> [1] "weblm"

   str(wordCandidates, max.level = 1)
   #> List of 3
   #>  $ results:'data.frame':  5 obs. of  2 variables:
   #>  $ json   : chr "{"candidates":[{"word":"doing","probability":-1.105}, __truncated__ ]}
   #>  $ request:List of 7
   #>   ..- attr(*, "class")= chr "request"
   #>  - attr(*, "class")= chr "weblm"

   # Print results
   pandoc.table(wordCandidates$results)
   #> ---------------------
   #>   word    probability
   #> ------- -------------
   #>   doing     -1.105
   #>
   #>    in       -1.239
   #>
   #> feeling     -1.249
   #>
   #>   going     -1.378
   #>
   #>   today      -1.43
   #> ---------------------

 }, error = function(err) {

   # Print error
   geterrmessage()

 })
## End(Not run)

Author(s)

Phil Ferriere pferriere@hotmail.com