test(bm25-vectorizer): add test cases

Author: sanjayaksaxena

test/bm25-vectorizer-specs.js

@@ -0,0 +1,306 @@
+//     wink-nlp
+//
+//     Copyright (C) GRAYPE Systems Private Limited
+//
+//     This file is part of “wink-nlp”.
+//
+//     Permission is hereby granted, free of charge, to any
+//     person obtaining a copy of this software and
+//     associated documentation files (the "Software"), to
+//     deal in the Software without restriction, including
+//     without limitation the rights to use, copy, modify,
+//     merge, publish, distribute, sublicense, and/or sell
+//     copies of the Software, and to permit persons to
+//     whom the Software is furnished to do so, subject to
+//     the following conditions:
+//
+//     The above copyright notice and this permission notice
+//     shall be included in all copies or substantial
+//     portions of the Software.
+//
+//     THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF
+//     ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED
+//     TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
+//     PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+//     THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
+//     DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF
+//     CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+//     CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+//     DEALINGS IN THE SOFTWARE.
+
+//
+
+var chai = require( 'chai' );
+var mocha = require( 'mocha' );
+var bm25 = require( '../utilities/bm25-vectorizer.js' );
+var its = require( '../src/its.js' );
+
+
+var expect = chai.expect;
+var describe = mocha.describe;
+var it = mocha.it;
+
+describe( 'bm25-vectorizer', function () {
+  describe( 'configurations', function () {
+    it( 'empty config should return default values', function () {
+      expect( bm25().config() ).to.deep.equal( { k: 1, k1: 1.2, b: 0.75, norm: 'none' } );
+    } );
+
+    it( 'null config should return default values', function () {
+      expect( bm25().config( { k: null, k1: null, b: null, norm: null } ) ).to.deep.equal( { k: 1, k1: 1.2, b: 0.75, norm: 'none' } );
+    } );
+
+    it( 'NaN/non-string config should return default values', function () {
+      expect( bm25( { k: [], k1: {}, b: 'null', norm: 'null' } ).config() ).to.deep.equal( { k: 1, k1: 1.2, b: 0.75, norm: 'none' } );
+    } );
+
+    it( '-ve values in config should return default values', function () {
+      expect( bm25( { k: -1, k1: -1, b: -1, norm: 'null' } ).config() ).to.deep.equal( { k: 1, k1: 1.2, b: 0.75, norm: 'none' } );
+    } );
+
+    it( 'large +ve values in config should return default values', function () {
+      expect( bm25( { k: 101, k1: 101, b: 2, norm: 'null' } ).config() ).to.deep.equal( { k: 1, k1: 1.2, b: 0.75, norm: 'none' } );
+    } );
+
+    it( 'valid values in config should return defined values', function () {
+      expect( bm25( { k: 2, k1: 1.5, b: 0.6, norm: 'l2' } ).config() ).to.deep.equal( { k: 2, k1: 1.5, b: 0.60, norm: 'l2' } );
+    } );
+  } );
+
+  describe( 'learn from 0-documents', function () {
+    const v = bm25();
+    it( '.out() should return []', function () {
+      expect( v.out() ).to.deep.equal( [] );
+    } );
+
+    it( '.out( its.docTermMatrix ) should return []', function () {
+      expect( v.out( its.docTermMatrix ) ).to.deep.equal( [] );
+    } );
+
+    it( '.out( its.docBOWArray ) should return []', function () {
+      expect( v.out( its.docBOWArray ) ).to.deep.equal( [] );
+    } );
+
+    it( '.out( its.terms ) should return []', function () {
+      expect( v.out( its.terms ) ).to.deep.equal( [] );
+    } );
+
+    it( '.out( its.idf ) should return []', function () {
+      expect( v.out( its.idf ) ).to.deep.equal( [] );
+    } );
+
+    it( '.out( its.modelJSON ) should return []', function () {
+      expect( v.out( its.modelJSON ) ).to.deep.equal( JSON.stringify( { tf: [], idf: {} } ) );
+    } );
+
+    it( '.length() should return []', function () {
+      expect( v.length() ).to.equal( 0 );
+    } );
+
+    it( 'doc.out() should return undefined', function () {
+      expect( v.doc( 0 ).out() ).to.deep.equal( undefined );
+    } );
+
+    it( 'doc.out( its.tf ) should return []', function () {
+      expect( v.doc( 0 ).out( its.tf ) ).to.deep.equal( [] );
+    } );
+
+    it( 'doc.out( its.vector ) should return []', function () {
+      expect( v.doc( 0 ).out( its.vector ) ).to.deep.equal( [] );
+    } );
+
+    it( 'doc.out( its.bow ) should return []', function () {
+      expect( v.doc( 0 ).out( its.bow ) ).to.deep.equal( undefined );
+    } );
+
+    it( 'doc.length() should return []', function () {
+      expect( v.doc( 0 ).length() ).to.equal( 0 );
+    } );
+
+    // it( 'should throw error if readDoc is given non-text', function () {
+    //   expect( nlp.readDoc.bind( 1 ) ).to.throw( /^wink-nlp: expecting a valid Javascript string/ );
+    // } );
+  } );
+
+  describe( 'learn from 1-document', function () {
+    const bow = { rain: 0.395562849, go: 0.287682072, away: 0.287682072 };
+    const json = '{"tf":[{"rain":0.395562849,"go":0.287682072,"away":0.287682072}],"idf":{"rain":0.287682072,"go":0.287682072,"away":0.287682072}}';
+    const v = bm25();
+    v.learn( 'rain rain go away'.split( /\s+/g ) );
+
+    it( 'out() should return [ bow ]', function () {
+      expect( v.out() ).to.deep.equal( [ bow ] );
+    } );
+
+    it( 'out( its.bow ) should return [ bow ]', function () {
+      expect( v.out( its.bow ) ).to.deep.equal( [ bow ] );
+    } );
+
+    it( 'out( its.docTermMatrix ) should return document term matrix', function () {
+      expect( v.out( its.docTermMatrix ) ).to.deep.equal( [ [ 0.287682072, 0.287682072, 0.395562849 ] ] );
+    } );
+
+    it( 'out( its.docBOWArray ) should return [ bow ]', function () {
+      expect( v.out( its.docBOWArray ) ).to.deep.equal( [ bow ] );
+    } );
+
+    it( 'out( its.idf ) should return [ bow ]', function () {
+      expect( v.out( its.idf ) ).to.deep.equal( [ [ 'away', 0.287682072 ], [ 'go', 0.287682072 ], [ 'rain', 0.287682072 ] ] );
+    } );
+
+    it( 'out( its.terms ) should return sorted array of terms', function () {
+      expect( v.out( its.terms ) ).to.deep.equal( [ 'away', 'go', 'rain' ] );
+    } );
+
+    it( 'out( its.modelJSON ) should return models JSON', function () {
+      expect( v.out( its.modelJSON ) ).to.deep.equal( json );
+    } );
+
+    it( 'vectorOf() should return vector of tokens', function () {
+      expect( v.vectorOf( [ 'rain', 'is', 'going', 'away' ] ) ).to.deep.equal( [ 0.287682072, 0, 0.287682072 ] );
+    } );
+
+    it( 'length() should return 3', function () {
+      expect( v.length() ).to.equal( 3 );
+    } );
+
+    it( 'doc.out( its.tf ) should return freq table of terms', function () {
+      expect( v.doc( 0 ).out( its.tf ) ).to.deep.equal( [ [ 'rain', 0.395562849 ], [ 'away', 0.287682072 ], [ 'go', 0.287682072 ] ] );
+    } );
+
+    it( 'doc.out( its.vector ) should return its vector', function () {
+      expect( v.doc( 0 ).out( its.vector ) ).to.deep.equal( [ 0.287682072, 0.287682072, 0.395562849 ] );
+    } );
+
+    it( 'doc.out( its.bow ) should return its bow', function () {
+      expect( v.doc( 0 ).out( its.bow ) ).to.deep.equal( bow );
+    } );
+
+    it( 'doc.length() should return 3', function () {
+      expect( v.doc( 0 ).length() ).to.equal( 3 );
+    } );
+  } );
+
+  describe( 'learn from multiple documents with l2 norm', function () {
+    const terms = [ 'are', 'black', 'blue', 'cats', 'rats', 'some', 'white' ];
+    const idf = [
+      [ 'black', 0.980829253 ],
+      [ 'blue', 0.980829253 ],
+      [ 'cats', 0.980829253 ],
+      [ 'some', 0.980829253 ],
+      [ 'white', 0.980829253 ],
+      [ 'rats', 0.470003629 ],
+      [ 'are', 0.133531393 ]
+    ];
+    const dtm = [
+      [ 0.121858341, 0, 0.895087087, 0, 0.428916835, 0, 0 ],
+      [ 0.095823468, 0, 0, 0.703852919, 0, 0, 0.703852919 ],
+      [ 0.086275085, 0.633717097, 0, 0, 0.435157318, 0.633717097, 0 ]
+    ];
+    const v = bm25( { norm: 'l2' } );
+    v.learn( 'rats are blue'.split( /\s+/g ) );
+    v.learn( 'cats are white'.split( /\s+/g ) );
+    v.learn( 'some rats rats are black'.split( /\s+/g ) );
+
+    it( 'doc.out( its.vector ) should return its vector', function () {
+      expect( v.doc( 2 ).out( its.vector ) ).to.deep.equal( [ 0.086275085, 0.633717097, 0, 0, 0.435157318, 0.633717097, 0 ] );
+    } );
+
+    it( 'out( its.idf ) should return its idfs freq table', function () {
+      expect( v.out( its.idf ) ).to.deep.equal( idf );
+    } );
+
+    it( 'out( its.terms ) should return its doc terms in alpha sort', function () {
+      expect( v.out( its.terms ) ).to.deep.equal( terms );
+    } );
+
+    it( 'out( its.docTermMatrix ) should return its doc term matrix', function () {
+      expect( v.out( its.docTermMatrix ) ).to.deep.equal( dtm );
+    } );
+
+    it( 'vectorOf() should return its vector', function () {
+      expect( v.vectorOf( 'rats were blue'.split( /\s+/g ) ) ).to.deep.equal( [ 0, 0, 0.901807807, 0, 0.432137338, 0, 0 ] );
+    } );
+  } );
+
+  describe( 'learn from multiple documents with l1 norm', function () {
+    const terms = [ 'are', 'black', 'blue', 'cats', 'rats', 'some', 'white' ];
+    const idf = [
+      [ 'black', 0.980829253 ],
+      [ 'blue', 0.980829253 ],
+      [ 'cats', 0.980829253 ],
+      [ 'some', 0.980829253 ],
+      [ 'white', 0.980829253 ],
+      [ 'rats', 0.470003629 ],
+      [ 'are', 0.133531393 ]
+    ];
+    const dtm = [
+      [ 0.08428074, 0, 0.619068019, 0, 0.296651241, 0, 0 ],
+      [ 0.063732358, 0, 0, 0.468133821, 0, 0, 0.468133821 ],
+      [ 0.048228909, 0.354256208, 0, 0, 0.243258675, 0.354256208, 0 ]
+    ];
+    const v = bm25( { norm: 'l1' } );
+    v.learn( 'rats are blue'.split( /\s+/g ) );
+    v.learn( 'cats are white'.split( /\s+/g ) );
+    v.learn( 'some rats rats are black'.split( /\s+/g ) );
+
+    it( 'doc.out( its.vector ) should return its vector', function () {
+      expect( v.doc( 2 ).out( its.vector ) ).to.deep.equal( [ 0.048228909, 0.354256208, 0, 0, 0.243258675, 0.354256208, 0 ] );
+    } );
+
+    it( 'out( its.idf ) should return its idfs freq table', function () {
+      expect( v.out( its.idf ) ).to.deep.equal( idf );
+    } );
+
+    it( 'out( its.terms ) should return its doc terms in alpha sort', function () {
+      expect( v.out( its.terms ) ).to.deep.equal( terms );
+    } );
+
+    it( 'out( its.docTermMatrix ) should return its doc term matrix', function () {
+      expect( v.out( its.docTermMatrix ) ).to.deep.equal( dtm );
+    } );
+
+    it( 'should throw error learn() is called after out()', function () {
+      expect( v.learn.bind( [ 'hello', 'world' ] ) ).to.throw( 'wink-nlp: learn can not be used after a call to out() API in BM25 Vectorizer' );
+    } );
+  } );
+
+  describe( 'values of TF & IDF', function () {
+    const v = bm25( { norm: 'l1' } );
+    // johann: `ln( 1 + ( ( 4 - 2 + 0.5 ) / ( 2 + 0.5 ) ) ) = 0.693147181`
+    // bach: `ln( 1 + ( ( 4 - 4 + 0.5 ) / ( 4 + 0.5 ) ) ) = 0.105360516`
+    const model = '{"tf":[{"bach":1},{"j":0.919531173,"bach":0.080468827},{"johann":0.346144285,"s":0.601240713,"bach":0.052615002},{"johann":0.346144285,"sebastian":0.601240713,"bach":0.052615002}],"idf":{"bach":0.105360516,"j":1.203972804,"johann":0.693147181,"s":1.203972804,"sebastian":1.203972804}}';
+    v.learn( 'Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'J Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'Johann S Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'Johann Sebastian Bach'.toLowerCase().split( /\s+/g ) );
+
+    it( 'should return correct idf values', function () {
+      expect( v.out( its.modelJSON ) ).to.equal( model );
+    } );
+  } );
+
+  describe( 'completely OOV tokens with l1 norm', function () {
+    const v = bm25( { norm: 'l1' } );
+    v.learn( 'Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'J Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'Johann S Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'Johann Sebastian Bach'.toLowerCase().split( /\s+/g ) );
+
+    it( 'should return 0-vector', function () {
+      expect( v.vectorOf([ 'cat', 'cat', 'green', 'is' ] ) ).to.deep.equal( [ 0, 0, 0, 0, 0 ] );
+    } );
+  } );
+
+  describe( 'completely OOV tokens with l2 norm', function () {
+    const v = bm25( { norm: 'l2' } );
+    v.learn( 'Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'J Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'Johann S Bach'.toLowerCase().split( /\s+/g ) );
+    v.learn( 'Johann Sebastian Bach'.toLowerCase().split( /\s+/g ) );
+
+    it( 'should return 0-vector', function () {
+      expect( v.vectorOf([ 'cat', 'cat', 'green', 'is' ] ) ).to.deep.equal( [ 0, 0, 0, 0, 0 ] );
+    } );
+  } );
+} );