{mscsweblm4r} vignette

What’s the Web LM REST API?

Microsoft Cognitive Services – formerly known as Project Oxford – are a set of APIs, SDKs and services that developers can use to add AI features to their apps. Those features include emotion and video detection; facial, speech and vision recognition; and speech and language understanding.

The Web Language Model REST API provides tools for natural language processing NLP.

Per Microsoft’s website, this API uses smoothed Backoff N-gram language models (supporting Markov order up to 5) that were trained on four web-scale American English corpora collected by Bing (web page body, title, anchor and query).

The MSCS Web LM REST API supports the following lookup operations:

Calculate the joint probability that a sequence of words will appear together.
Compute the conditional probability that a specific word will follow an existing sequence of words.
Get the list of words (completions) most likely to follow a given sequence of words.
Insert spaces into a string of words adjoined together without any spaces (hashtags, URLs, etc.).
Retrieve the list of supported language models.

Package Installation

You can either install the latest stable version from CRAN:

if ("mscsweblm4r" %in% installed.packages()[,"Package"] == FALSE) {
  install.packages("mscsweblm4r")
}

Or, you can install the development version

if ("mscsweblm4r" %in% installed.packages()[,"Package"] == FALSE) {
  if ("devtools" %in% installed.packages()[,"Package"] == FALSE) {
    install.packages("devtools")
  }
  devtools::install_github("philferriere/mscsweblm4r")
}

Package Loading and Configuration:

After loading {mscsweblm4r} with library(), you must call weblmInit() before you can call any of the core {mscsweblm4r} functions.

The weblmInit() configuration function will first check to see if the variable MSCS_WEBLANGUAGEMODEL_CONFIG_FILE exists in the system environment. If it does, the package will use that as the path to the configuration file.

If MSCS_WEBLANGUAGEMODEL_CONFIG_FILE doesn’t exist, it will look for the file .mscskeys.json in the current user’s home directory (that’s ~/.mscskeys.json on Linux, and something like C:\Users\Phil\Documents\.mscskeys.json on Windows). If the file is found, the package will load the API key and URL from it.

If using a file, please make sure it has the following structure:

{
  "weblanguagemodelurl": "https://api.projectoxford.ai/text/weblm/v1.0/",
  "weblanguagemodelkey": "...MSCS Web Language Model API key goes here..."
}

If no configuration file is found, weblmInit() will attempt to pick up its configuration from two Sys env variables instead:

MSCS_WEBLANGUAGEMODEL_URL - the URL for the Web LM REST API.

MSCS_WEBLANGUAGEMODEL_KEY - your personal Web LM REST API key.

weblmInit() needs to be called only once, after package load.

Error Handling

The MSCS Web LM API is a RESTful API. HTTP requests over a network and the Internet can fail. Because of congestion, because the web site is down for maintenance, because of firewall configuration issues, etc. There are many possible points of failure.

The API can also fail if you’ve exhausted your call volume quota or are exceeding the API calls rate limit. Unfortunately, MSCS does not expose an API you can query to check if you’re about to exceed your quota for instance. The only way you’ll know for sure is by looking at the error code returned after an API call has failed.

Therefore, you must write your R code with failure in mind. Our preferred way is to use tryCatch(). Its mechanism may appear a bit daunting at first, but it is well documented. We’ve also included many examples, as you’ll see below.

Package Configuration with Error Handling

Here’s some sample code that illustrates how to use tryCatch():

library('mscsweblm4r')
tryCatch({

  weblmInit()

}, error = function(err) {

  geterrmessage()

})

If {mscsweblm4r} cannot locate .mscskeys.json nor any of the configuration environment variables, the code above will generate the following output:

[1] "mscsweblm4r: could not load config info from Sys env nor from file"

Similarly, weblmInit() will fail if {mscsweblm4r} cannot find the weblanguagemodelkey key in .mscskeys.json, or fails to parse it correctly, etc. This is why it is so important to use tryCatch() with all {mscsweblm4r} functions.

Package API

The five API calls exposed by {mscsweblm4r} are the following:

  # Retrieve a list of supported web language models
  weblmListAvailableModels()

  # Break a string of concatenated words into individual words
  weblmBreakIntoWords(
    textToBreak,                      # ASCII only
    modelToUse = "body",              # "title"|"anchor"|"query"(default)|"body"
    orderOfNgram = 5L,                # 1L|2L|3L|4L|5L(default)
    maxNumOfCandidatesReturned = 5L   # Default: 5L
  )

  # Get the words most likely to follow a sequence of words
  weblmGenerateNextWords(
    precedingWords,                  # ASCII only
    modelToUse = "title",            # "title"|"anchor"|"query"(default)|"body"
    orderOfNgram = 4L,               # 1L|2L|3L|4L|5L(default)
    maxNumOfCandidatesReturned = 5L  # Default: 5L
  )

  # Calculate joint probability a particular sequence of words will appear together
  weblmCalculateJointProbability(
    inputWords =,                    # ASCII only
    modelToUse = "query",            # "title"|"anchor"|"query"(default)|"body"
    orderOfNgram = 4L                # 1L|2L|3L|4L|5L(default)
  )

  # Calculate conditional probability a particular word will follow a given sequence of words
  weblmCalculateConditionalProbability(
    precedingWords,                  # ASCII only
    continuations,                   # ASCII only
    modelToUse = "title",            # "title"|"anchor"|"query"(default)|"body"
    orderOfNgram = 4L                # 1L|2L|3L|4L|5L(default)
  )

These functions return S3 class objects of the class weblm. The weblm object exposes formatted results (in data.frame format), the REST API JSON response (should you care), and the HTTP request (mostly for debugging purposes).

Sample Code

The following code snippets illustrate how to use {mscsweblm4r} functions and show what results they return with toy examples. If after reviewing this code there is still confusion regarding how and when to use each function, please refer to the original documentation.

List Available Models function

tryCatch({

  # Retrieve a list of supported web language models
  weblmListAvailableModels()

}, error = function(err) {

 # Print error
 geterrmessage()

})

Break Into Words function

tryCatch({

  # Break a sentence into words
  weblmBreakIntoWords(
    textToBreak = "testforwordbreak", # ASCII only
    modelToUse = "body",              # "title"|"anchor"|"query"(default)|"body"
    orderOfNgram = 5L,                # 1L|2L|3L|4L|5L(default)
    maxNumOfCandidatesReturned = 5L   # Default: 5L
  )

}, error = function(err) {

  # Print error
  geterrmessage()

})

Generate Next Word function

tryCatch({

  # Generate next words
  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
  )

}, error = function(err) {

  # Print error
  geterrmessage()

})

Calculate Joint Probability function

tryCatch({

  # Calculate joint probability a particular sequence of words will appear together
  weblmCalculateJointProbability(
    inputWords = c("where", "is", "San", "Francisco", "where is",
                   "San Francisco", "where is San Francisco"),  # ASCII only
    modelToUse = "query",                     # "title"|"anchor"|"query"(default)|"body"
    orderOfNgram = 4L                         # 1L|2L|3L|4L|5L(default)
  )

}, error = function(err) {

  # Print error
  geterrmessage()

})

Calculate Conditional Probability function

tryCatch({

  # Calculate conditional probability a particular word will follow a given sequence of words
  weblmCalculateConditionalProbability(
    precedingWords = "hello world wide",       # ASCII only
    continuations = c("web", "range", "open"), # ASCII only
    modelToUse = "title",                      # "title"|"anchor"|"query"(default)|"body"
    orderOfNgram = 4L                          # 1L|2L|3L|4L|5L(default)
  )

}, error = function(err) {

  # Print error
  geterrmessage()

})

Demo Application

A test/demo Shiny web application is available here