Click here if you are not redirected.

A to Z, F15

N-Grams and Markov Chains


Aside examples (many callbacks, many DOM elements)

Markov projects

Exercise ideas

N-Grams and Markov Chains

The infinite monkey theorem suggests that a monkey randomly typing at a keyboard for an infinite amount of time would eventually type the complete works of William Shakespeare. Working out the math to this problem reveals just how insanely long this would actually take. Even for a monkey to randomly type “to be or not to be” requires eons upon eons of time. We can demonstrate the idea with JavaScript by generating random Strings.

// All of our possible characters
var possible = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOP .';
// The random String
var tobe = '';
// Pick 18 random characters
for (var i = 0; i < 18; i++) {
 var r = possible.charAt(floor(random(0,possible.length)));
 tobe += r;

Have you seen “to be or not to be” yet? Probably not. After all, a true monkey types totally at random, meaning there is an equal probability that any key will be pressed at any given time. The chances are incredibly slim we’ll get an exact sequence. But what if we could generate our own custom probability map for a keyboard (or a sequence of word tokens)? What kind of results could we achieve? Let’s consider the following scenario. What if we picked letters according to their actual frequency in the English language? Here’s a chart to get us started.

English letter frequency (alphabetic).svg
English letter frequency (alphabetic) by Nandhp Licensed under Public domain via Wikimedia Commons.

Letter Probability    
------ -----------
a      8.167%         
b      1.492%         
c      2.782%         
d      4.253%        

How might we pick ‘a’ 8% of the time and ‘d’ 4% of the time? Another way of stating this question might be: “How would be pick ‘a’ twice as often as ‘d’”? While there are a variety of solutions to this sort of problem a simple one that we will employ in our examples is the following. Imagine we had an array.

// An array of two choices
var possibilities = ['a','d'];

We could then pick one element from that array randomly as follows:

// A random integer from 0 to the end of the array
var i = floor(random(possibilities.length));
// Get the element in that index
var pick = possibilities[i];

Each element in the case of an array with two spots has a 50% chance of being picked. But what if created the array as follows instead?

// Two choices, but one appears twice
var possibilities = ['a','a', 'd'];

Now when picking a random element, there is a two out of three chance of picking ‘a’. In fact, ‘a’ is twice as likely to be picked than ‘d’. Using all the letter frequencies found in this JSON file, we could rewrite our random generator to build an array of letters, adding each letter to the array a number of times according to its frequency.

// All the possible letters
var letters = 'abcdefghijklmnopqrstuvwxyz';
for (var i = 0; i < letters.length; i++) {
    // Let's assume we've loaded a JSON object
    // that has the probabilities for each one
    var prob = json[letters[i]];
    // Arbitrarily multiplying the probability by 1,000
    // If the probability is 1.067% then it would be
    // placed in the array 1,067 times.
    var n = floor(prob*1000);
    for (var j = 0; j < n; j++) {

While we’ve probably increased the likelihood of randomly typing “to be or not to be” only slightly, we can nevertheless see how the quality of the results have changed. e’s appear a great deal more than z’s and so on and so forth.

Next, let’s take this idea of custom letter probabilities another step forward and examine the frequencies of letters neighboring other letters commonly in English. For example, how often does ‘h’ occur after ‘t’ versus ‘a’ or ‘r’ and so on and so forth?

This is exactly the approach of N-grams. An n-gram is a contiguous sequence of N elements. In the case of text, this might be N letters or N words or N phonemes or N syllables. For an example, here’s a function that returns all word N-grams for a given String (using a regex to split the text up into tokens):

function nGrams(sentence, n) {
  // Split sentence up into words
  var words = sentence.split(/\W+/);
  // Total number of n-grams we will have
  var total = words.length - n;
  var grams = [];
  // Loop through and create all sequences
  for(var i = 0; i <= total; i++) {
    var seq = '';
    for (var j = i; j < i + n; j++) {
      seq += words[j] + ' ';
    // Add to array
  return grams;

Try it out below.


Looking at all the N-grams of a given text provides a strategy for generating text. Let’s go back to the phrase “to_be_or_not_to_be” (using an underscore instead of space for clarity). Let’s start with the simplest possible scenario and look at all N-grams where N=1 (which is admittedly a bit silly, but will be a useful demonstration.)

t o _ b e _ o r _ n o t _ t o _ b e

Now, let’s look at all the possible characters that appear after t:

t -> o
t -> _
t -> o

From the above we can calculate that in this (very small) piece of text, an ‘o’ follows a ‘t’ 67% of the time and a space 33%. Now let’s look at ‘o’, where we next see a space 50% of the time, ‘r’ 25%, and ‘t’ 25%.

o -> _
o -> r
o -> t
o -> _

We could express this result as a JavaScript object:

// All ngrams and next characters
var ngrams = {
  't': ['o', '_', 'o'],
  'o': ['_', 'r', 't', '_']

Now, imagine we next decided to generate new text based on these probabilities. We might start with ‘t’:

var txt = 't';

And then pick a new character based on the array associated in the ngrams object.

var txt = 't';

var possible = ngrams['t'];
var r = floor(random(possible.length));
var next = possible[r];

txt = txt + next;

Now repeat for the next N-gram. And so and and so forth. Here are some sample outcomes (and the full code for producing these):

// Our simple ngrams model
var ngrams = {
  't': ['o', '_', 'o'],
  'o': ['_', 'r', 't', '_']

// Starting text
var txt = 't';
// Let's do the following ten times
for (var i = 0; i < 10; i++) {
  var possible = ngrams[txt.charAt(i)];
  // If there is nothing in the ngrams object for this character
  // it's a "terminal" character and we stop
  if (!possible) {
  // Otherwise pick a random number
  var r = floor(random(possible.length));
  // And we've got the next chararacter
  var next = possible[r];

This technique is known as a Markov Chain. A Markov Chain can be described as a sequence of random “states” where each new state is conditional only on the previous state. An example of a Markov Chain is monopoly. The next state of the monopoly board depends on the current state and the roll of the dice. It doesn’t matter how we got to that current state, only what it is at the moment. A game like blackjack, for example, is different in that the deal of the cards is dependent on the history of many previous deals (assuming a single-deck not continuously shuffled.)

Using an N-gram model, can use a markov chain to generate text where each new word or character is dependent on the previous word (or character) or sequence of words (or characters). For example. given the phrase “I have to” we might say the next word is 50% likely to be “go”, 30% likely to be “run” and 20% likely to be “pee.” We can construct these word sequence probabilities based on a large corpus of source texts. Here, for example, is the full set of ngrams of order 2 and their possible outcomes in the phrase: “to be or not to be, that is the question.”

var ngrams = {
   "to": [ " ", " " ],
   "o ": [ "b", "b" ],
   " b": [ "e", "e" ],
   "be": [ " ", "," ],
   "e ": [ "o", "q" ],
   " o": [ "r" ],
   "or": [ " " ],
   "r ": [ "n" ],
   " n": [ "o" ],
   "no": [ "t" ],
   "ot": [ " " ],
   "t ": [ "t", "i" ],
   " t": [ "o", "h", "h" ],
   "e,": [ " " ],
   ", ": [ "t" ],
   "th": [ "a", "e" ],
   "ha": [ "t" ],
   "at": [ " " ],
   " i": [ "s" ],
   "is": [ " " ],
   "s ": [ "t" ],
   "he": [ " " ],
   " q": [ "u" ],
   "qu": [ "e" ],
   "ue": [ "s" ],
   "es": [ "t" ],
   "st": [ "i" ],
   "ti": [ "o" ],
   "io": [ "n" ],
   "on": [ "." ]

The process to generate the above ngrams object is quite similar to the concordance we previously developed. Only this time, instead of pairing each token with a count, we are pairing each N-gram with an array of possible next characters. Let’s look at how this is built.

var ngrams = {};
var text = 'to be or not to be, that is the question';
var n = 2;
// Look at all characters of the String
for (var i = 0; i < text.length - n; i++) {
  // Look at an ngram
  var gram = text.substring(i, i + n);
  // Look at the next charaacter
  var next = text.charAt(i + n);
  // Is this a new one, make an empty array
  if (!ngrams.hasOwnProperty(gram)) {
    ngrams[gram] = [];
  // Add the next character as a possible outcome

A small difference from the concordance you might notice above is the use of hasOwnProperty() which is a universal object method in JavaScript that allows you to test if a variable is a property of the object or not. Last week, we said if (ngrams[gram] === undefined). Both are valid strategies.

Once we’ve got the ngrams object with all possibities mapped out we can start to generate text.

// Start with an arbitrary ngram
var current = 'to';
// The final text
var output = current;

// Generate a new character some number of times
for (var i = 0; i < 20; i++) {
  // If this is a valid ngram
  if (ngrams.hasOwnProperty(current)) {
    // What are all the possible next tokens
    var possible = ngrams[current];
    // Pick one randomly
    var next = choice(possible);
    // Add to the output
    output += next;
    // Get the last N entries of the output; we'll use this to look up
    // an ngram in the next iteration of the loop
    current = output.substring(output.length-n, output.length);
  } else {

You might notice the choice() function above. Just about all of our examples require the ability to pick a random element from an array over and over. To simplify this process, I’m emulating the python choice() function.

function choice(somelist) {
  var i = floor(random(somelist.length));
  return somelist[i];

In the examples, you’ll find all of the above packaged into a MarkovGenerator object. Special thanks to Allison Parish’s excellent RWET examples which this is modeled after.