node package manager

word2vector

word2vector NodeJS Interface

============= This is a Node.js interface for Google's word2vector.
Here is an example of how to load large model like GoogleNews-vectors-negative300.bin by this package.
# Supports both binary model and raw text model. # Warning: Windows is not supported.

Installation

Linux, Unix OS are supported. Install it via npm:

npm install word2vector --save

In Node.js, require the module as below:

var w2v = require( 'word2vector' );

API Document:


Overview

train load getVector getVectors getSimilarWords getNeighbors similarity substract add


w2v.train( trainFile, modelFile, options, callback )

Click here to see example TrainFile format.
Example:

var w2v = require("./lib");
var trainFile = "./data/train.data",
    modelFile = "./data/test.model.bin";
w2v.train(trainFile, modelFile, {
   cbow: 1,           // use the continuous bag of words model //default 
   size: 10,          // sets the size (dimension) of word vectors // default 100 
   window: 8,         // sets maximal skip length between words // default 5 
    binary: 1,         // save the resulting vectors in binary mode // default off 
   negative: 25,      // number of negative examples; common values are 3 - 10 (0 = not used) // default 5 
   hs: 0,             // 1 = use  Hierarchical Softmax // default 0 
   sample: 1e-4,      
   threads: 20,
   iter: 15,
   minCount: 1,       // This will discard words that appear less than *minCount* times // default 5 
    logOn: false       // sets whether any output should be printed to the console // default false 
  });

w2v.load( modelFile,?readType = "")

Should load model before call any calcuation functions.
Params Description Default Value
readType Model format, pass "utf-8" if using a raw text model. "bin"

Example:

var w2v = require("../lib");
var modelFile = "./test.model.bin";
w2v.load( modelFile );
// console.log(w2v.getSimilarWordsWords()); 

w2v.getVector(word="word")

Params Description Default Value
word String to be searched. "word"

Example:

'use strict';
var w2v = require("./lib");
var modelFile = "./data/test.model.bin";
w2v.load( modelFile );
console.log(w2v.getVector("孫悟空"));
console.log(w2v.getVector("李洵"));

Sample Output:

// Array Type Only 
[ 0.104406,
  -0.160019,
  -0.604506,
  -0.622804,
  0.039482,
  -0.120058,
  0.073555,
  0.05646,
  0.099059,
  -0.419282 ]
 
null // Return null if this word is not in model. 

w2v.getVectors(words=["word1", "word2"], ?options = {})

Params Description Default Value
words Array of strings to be searched. "word"
options.returnType return Array or Object type Array

Example:

var w2v = require("./lib");  
var modelFile = "./data/test.model.bin";
w2v.load( modelFile );
console.log(w2v.getVectors(["孫悟空", "李洵"]));
console.log(w2v.getVectors(["孫悟空", "李洵"], {returnType: "Object"}));

Sample Output:

// Array Type 
[ { word: '孫悟空',
    vector:
     [ 0.104406,
       -0.160019,
       -0.604506,
       -0.622804,
       0.039482,
       -0.120058,
       0.073555,
       0.05646,
       0.099059,
       -0.419282 ] },
  { word: '李洵', vector: null } ]
  // this will trigger a error log in console: 
  //'李洵' is not found in the model. 
// Object Type 
{ '孫悟空':
   [ 0.104406,
     -0.160019,
     -0.604506,
     -0.622804,
     0.039482,
     -0.120058,
     0.073555,
     0.05646,
     0.099059,
     -0.419282 ],
  '李洵': null }
  // this will trigger a error log in console: 
  //'李洵' is not found in the model. 

w2v.getSimilarWords(word = "word", ?options = {})

Return 40ish words that is similar to "word".
Params Description Default Value
word Strings to be searched. "word"
options.N return topN results Array
options.returnType return Array or Object type Array

Example:

var w2v = require("./lib");
var modelFile = "./data/test.model.bin";
w2v.load( modelFile );
console.log(w2v.getSimilarWords("唐三藏"));
console.log(w2v.getSimilarWords("李洵"));
console.log(w2v.getSimilarWords("唐三藏", {N:10, returnType: "Object"}));
console.log(w2v.getSimilarWords("李洵", {N:10, returnType: "Object"}));

Sample Output:

// Array Type 
[ { word: '孫悟空', cosineDistance: 0.974369 },
  { word: '吳承恩', cosineDistance: 0.96686 },
  { word: '林黛玉', cosineDistance: 0.966664 },
  { word: '北地', cosineDistance: 0.96264 },
  { word: '賈寶玉', cosineDistance: 0.962137 },
  { word: '楚霸王', cosineDistance: 0.955795 },
  { word: '梁山泊', cosineDistance: 0.932804 },
  { word: '濮陽', cosineDistance: 0.927542 },
  { word: '黃天霸', cosineDistance: 0.927459 },
  { word: '英雄豪傑', cosineDistance: 0.921575 }]
// Return empty [] if this word is not in model. 
'李洵' is not found in the model.
[]
// Object Type 
{ '孫悟空': 0.974369,
  '吳承恩': 0.96686,
  '林黛玉': 0.966664,
  '北地': 0.96264,
  '賈寶玉': 0.962137,
  '楚霸王': 0.955795,
  '梁山泊': 0.932804,
  '濮陽': 0.927542,
  '黃天霸': 0.927459,
  '英雄豪傑': 0.921575}
// Return empty {} if this word is not in model. 
'李洵' is not found in the model.
{}

w2v.getSimilarAsync(word = "word", options, callback)

...........................Discarded.................................

getNeighbors(vector, ?options = {})

Params Description Default Value
vector Vector to be searched. "word"
options.N return topN results Array
options.returnType return Array or Object type Array

Example1:

var w2v = require("./lib");
var modelFile = "./data/test.model.bin";
w2v.load( modelFile );
var a = w2v.getNeighbors(w2v.getVector("唐三藏"), {N: 9});
var b = w2v.getNeighbors(w2v.getVector("唐三藏"), {N: 9, returnType: "Object"});
// These are equal to use w2v.getSimilarWords("唐三藏"); 
console.log(a);
console.log(b);

Sample Output1:

[ { word: '唐三藏', cosineDistance: 0.9999993515200001 },
  { word: '孫悟空', cosineDistance: 0.974368825898 },
  { word: '吳承恩', cosineDistance: 0.966859435824 },
  { word: '林黛玉', cosineDistance: 0.966663471323 },
  { word: '北地', cosineDistance: 0.962639240211 },
  { word: '賈寶玉', cosineDistance: 0.9621371820049999 },
  { word: '楚霸王', cosineDistance: 0.9557946924850002 },
  { word: '梁山泊', cosineDistance: 0.9328033548890001 },
  { word: '濮陽', cosineDistance: 0.9275417727409999 } ]
{ '唐三藏': 0.9999993515200001,
  '孫悟空': 0.974368825898,
  '吳承恩': 0.966859435824,
  '林黛玉': 0.966663471323,
  '北地': 0.962639240211,
  '賈寶玉': 0.9621371820049999,
  '楚霸王': 0.9557946924850002,
  '梁山泊': 0.9328033548890001,
  '濮陽': 0.9275417727409999 }

Example2:

var w2v = require("../lib");
var modelFile = "./test.model.bin";
w2v.load( modelFile );
var a = w2v.getVectors(['唐三藏'], {returnType: "Object"})['唐三藏'];
var b = w2v.getVectors(['孫悟空'], {returnType: "Object"})['孫悟空'];
var c = [];
for(var i=0; i<a.length; i++) c.push(a[i] - b[i]);
var d = w2v.getNeighbors(c);
console.log(d);
// vector can do substractioin, while this didn't  mean anything. But you can create a vector by yourself. 

Sample Output2:

[ { word: '蒙上帝', cosineDistance: 0.794216 },
  { word: '阿房宮賦', cosineDistance: 0.787006 },
  { word: '玄秘', cosineDistance: 0.770159 },
  { word: '檀香山', cosineDistance: 0.755662 },
  { word: '先賢祠', cosineDistance: 0.746278 },
  { word: '蘇萊曼', cosineDistance: 0.745826 },
  { word: '盧梭', cosineDistance: 0.704465 },
  { word: '夏威夷', cosineDistance: 0.700885 },
  { word: '伏爾泰', cosineDistance: 0.698588 },
  { word: '杜爾哥', cosineDistance: 0.688763 },
  { word: '祝你們', cosineDistance: 0.687257 } ... ...],

w2v.similarity(word1 = "word1", word2 = "word2")

w2v.similarity(vector1 = [], word2 = "word2")

w2v.similarity(word1 = "word1", vector2 = [])

w2v.similarity(vector1 = [], vector2 = [])

Compute the [cosine similarity](https://en.wikipedia.org/wiki/Cosine_similarity) between the two vector.
Will auto search the vector of passed word in model. Return false if it's not found.
Params Description Default Value
word1 First Strings to be compared. No default value
word2 Second Strings to be compared. No default value
vector1 First Vector to be compared. No default value
vector2 Second Vector to be compared. No default value

Example:

'use strict';
var w2v = require("./lib");
var modelFile = "./data/test.model.bin";
w2v.load( modelFile );
var a = w2v.similarity("唐三藏", "孫悟空"); //  0.974368825898 
console.log(a);
var b = w2v.similarity("唐三藏", "李洵"); //  0.974368825898 
// same as var b = w2v.similarity("唐三藏", w2v.getVector("李洵")); 
// same as var b = w2v.similarity(w2v.getVector("唐三藏"), "李洵"); 
// same as var b = w2v.similarity(w2v.getVector("唐三藏"), w2v.getVector("李洵")); 
console.log(b);

Sample Output:

0.974368825898
// '李洵' is not found in the model. // error alert in console 
false

w2v.substract(word1 = "word1", word2 = "word2")

w2v.substract(vector1 = [], word2 = "word2")

w2v.substract(word1 = "word1", vector2 = [])

w2v.substract(vector1 = [], vector2 = [])

Substract vector1 from vector2.
Will auto search the vector of passed word in model. Return false if it's not found.
Params Description Default Value
word1 Subtrahend No default value
word2 Minuend No default value

Example:

'use strict';
var w2v = require("./lib");
var modelFile = "./data/test.model.bin";
w2v.load( modelFile );
var a = w2v.substract("孫悟空", "孫悟空");
console.log(a);

Sample Output:

[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]

w2v.add(word1 = "word1", word2 = "word2")

w2v.add(vector1 = [], word2 = "word2")

w2v.add(word1 = "word1", vector2 = [])

w2v.add(vector1 = [], vector2 = [])

Add vector1 to vector2.
Will auto search the vector of passed word in model. Return false if it's not found.
Params Description Default Value
word1 Summand No default value
word2 Addend No default value

Example:

'use strict';
var w2v = require("./lib");
var modelFile = "./data/test.model.bin";
w2v.load( modelFile );
var a = w2v.add("孫悟空", "孫悟空");
var b = w2v.getVector("孫悟空");
console.log(a);
console.log(b);

Sample Output:

[ 0.208812,
  -0.320038,
  -1.209012,
  -1.245608,
  0.078964,
  -0.240116,
  0.14711,
  0.11292,
  0.198118,
  -0.838564 ]
[ 0.104406,
  -0.160019,
  -0.604506,
  -0.622804,
  0.039482,
  -0.120058,
  0.073555,
  0.05646,
  0.099059,
  -0.419282 ]