nlptoolkit-postagger 1.0.0

package/dist/DummyPosTagger.d.ts

@@ -0,0 +1,20 @@
+import { PosTagger } from "./PosTagger";
+import { Sentence } from "nlptoolkit-corpus/dist/Sentence";
+import { PosTaggedCorpus } from "./PosTaggedCorpus";
+export declare class DummyPosTagger implements PosTagger {
+    private tagList;
+    /**
+     * Test method for the Dummy pos tagger. For each word, the method chooses randomly a tag from all possible
+     * tag list.
+     *
+     * @param sentence Sentence to be tagged.
+     * @return Annotated (tagged) sentence.
+     */
+    posTag(sentence: Sentence): Sentence;
+    /**
+     * Train method for the Dummy pos tagger. The algorithm gets all possible tag list.
+     *
+     * @param corpus Training data for the tagger.
+     */
+    train(corpus: PosTaggedCorpus): void;
+}

package/dist/DummyPosTagger.js

@@ -0,0 +1,47 @@
+(function (factory) {
+    if (typeof module === "object" && typeof module.exports === "object") {
+        var v = factory(require, exports);
+        if (v !== undefined) module.exports = v;
+    }
+    else if (typeof define === "function" && define.amd) {
+        define(["require", "exports", "nlptoolkit-corpus/dist/Sentence", "./PosTaggedWord"], factory);
+    }
+})(function (require, exports) {
+    "use strict";
+    Object.defineProperty(exports, "__esModule", { value: true });
+    exports.DummyPosTagger = void 0;
+    const Sentence_1 = require("nlptoolkit-corpus/dist/Sentence");
+    const PosTaggedWord_1 = require("./PosTaggedWord");
+    class DummyPosTagger {
+        constructor() {
+            this.tagList = new Array();
+        }
+        /**
+         * Test method for the Dummy pos tagger. For each word, the method chooses randomly a tag from all possible
+         * tag list.
+         *
+         * @param sentence Sentence to be tagged.
+         * @return Annotated (tagged) sentence.
+         */
+        posTag(sentence) {
+            let result = new Sentence_1.Sentence();
+            for (let i = 0; i < sentence.wordCount(); i++) {
+                result.addWord(new PosTaggedWord_1.PosTaggedWord(sentence.getWord(i).getName(), this.tagList[Math.floor(Math.random() * this.tagList.length)]));
+            }
+            return result;
+        }
+        /**
+         * Train method for the Dummy pos tagger. The algorithm gets all possible tag list.
+         *
+         * @param corpus Training data for the tagger.
+         */
+        train(corpus) {
+            let corpusTagList = corpus.getTagList();
+            for (let tag of corpusTagList) {
+                this.tagList.push(tag);
+            }
+        }
+    }
+    exports.DummyPosTagger = DummyPosTagger;
+});
+//# sourceMappingURL=DummyPosTagger.js.map

package/dist/DummyPosTagger.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"DummyPosTagger.js","sourceRoot":"","sources":["../source/DummyPosTagger.ts"],"names":[],"mappings":";;;;;;;;;;;;IACA,8DAAyD;IAEzD,mDAA8C;IAE9C,MAAa,cAAc;QAA3B;YAEY,YAAO,GAAkB,IAAI,KAAK,EAAU,CAAA;QA8BxD,CAAC;QA5BG;;;;;;WAMG;QACH,MAAM,CAAC,QAAkB;YACrB,IAAI,MAAM,GAAG,IAAI,mBAAQ,EAAE,CAAC;YAC5B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,SAAS,EAAE,EAAE,CAAC,EAAE,EAAC;gBAC1C,MAAM,CAAC,OAAO,CAAC,IAAI,6BAAa,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,OAAO,EAAE,EAC1D,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,MAAM,EAAE,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC;aACvE;YACD,OAAO,MAAM,CAAC;QAClB,CAAC;QAED;;;;WAIG;QACH,KAAK,CAAC,MAAuB;YACzB,IAAI,aAAa,GAAG,MAAM,CAAC,UAAU,EAAE,CAAC;YACxC,KAAK,IAAI,GAAG,IAAI,aAAa,EAAC;gBAC1B,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;aACzB;QACL,CAAC;KAEJ;IAhCD,wCAgCC"}

package/dist/HmmPosTagger.d.ts

@@ -0,0 +1,21 @@
+import { PosTagger } from "./PosTagger";
+import { Sentence } from "nlptoolkit-corpus/dist/Sentence";
+import { PosTaggedCorpus } from "./PosTaggedCorpus";
+export declare class HmmPosTagger implements PosTagger {
+    private hmm;
+    /**
+     * Test method for the Hmm pos tagger. For each sentence, the method uses the viterbi algorithm to produce the
+     * most possible state sequence for the given sentence.
+     *
+     * @param sentence Sentence to be tagged.
+     * @return Annotated (tagged) sentence.
+     */
+    posTag(sentence: Sentence): Sentence;
+    /**
+     * Train method for the Hmm pos tagger. The algorithm trains an Hmm from the corpus, where corpus constitutes
+     * as an observation array.
+     *
+     * @param corpus Training data for the tagger.
+     */
+    train(corpus: PosTaggedCorpus): void;
+}

package/dist/HmmPosTagger.js

@@ -0,0 +1,56 @@
+(function (factory) {
+    if (typeof module === "object" && typeof module.exports === "object") {
+        var v = factory(require, exports);
+        if (v !== undefined) module.exports = v;
+    }
+    else if (typeof define === "function" && define.amd) {
+        define(["require", "exports", "nlptoolkit-corpus/dist/Sentence", "./PosTaggedWord", "nlptoolkit-hmm/dist/Hmm1"], factory);
+    }
+})(function (require, exports) {
+    "use strict";
+    Object.defineProperty(exports, "__esModule", { value: true });
+    exports.HmmPosTagger = void 0;
+    const Sentence_1 = require("nlptoolkit-corpus/dist/Sentence");
+    const PosTaggedWord_1 = require("./PosTaggedWord");
+    const Hmm1_1 = require("nlptoolkit-hmm/dist/Hmm1");
+    class HmmPosTagger {
+        /**
+         * Test method for the Hmm pos tagger. For each sentence, the method uses the viterbi algorithm to produce the
+         * most possible state sequence for the given sentence.
+         *
+         * @param sentence Sentence to be tagged.
+         * @return Annotated (tagged) sentence.
+         */
+        posTag(sentence) {
+            let result = new Sentence_1.Sentence();
+            let tagList = this.hmm.viterbi(sentence.getWords());
+            for (let i = 0; i < sentence.wordCount(); i++) {
+                result.addWord(new PosTaggedWord_1.PosTaggedWord(sentence.getWord(i).getName(), tagList[i]));
+            }
+            return result;
+        }
+        /**
+         * Train method for the Hmm pos tagger. The algorithm trains an Hmm from the corpus, where corpus constitutes
+         * as an observation array.
+         *
+         * @param corpus Training data for the tagger.
+         */
+        train(corpus) {
+            let emittedSymbols = new Array();
+            for (let i = 0; i < emittedSymbols.length; i++) {
+                emittedSymbols.push(new Array());
+                for (let j = 0; j < corpus.getSentence(i).wordCount(); j++) {
+                    let word = corpus.getSentence(i).getWord(j);
+                    emittedSymbols[i].push(word.getTag());
+                }
+            }
+            let tagList = new Set();
+            for (let tag of corpus.getTagList()) {
+                tagList.add(tag);
+            }
+            this.hmm = new Hmm1_1.Hmm1(tagList, emittedSymbols, corpus.getAllWordsAsArrayList());
+        }
+    }
+    exports.HmmPosTagger = HmmPosTagger;
+});
+//# sourceMappingURL=HmmPosTagger.js.map

package/dist/HmmPosTagger.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"HmmPosTagger.js","sourceRoot":"","sources":["../source/HmmPosTagger.ts"],"names":[],"mappings":";;;;;;;;;;;;IACA,8DAAyD;IAIzD,mDAA8C;IAC9C,mDAA8C;IAE9C,MAAa,YAAY;QAIrB;;;;;;WAMG;QACH,MAAM,CAAC,QAAkB;YACrB,IAAI,MAAM,GAAG,IAAI,mBAAQ,EAAE,CAAC;YAC5B,IAAI,OAAO,GAAG,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,QAAQ,CAAC,QAAQ,EAAE,CAAC,CAAC;YACpD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,SAAS,EAAE,EAAE,CAAC,EAAE,EAAC;gBAC1C,MAAM,CAAC,OAAO,CAAC,IAAI,6BAAa,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,OAAO,EAAE,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;aAChF;YACD,OAAO,MAAM,CAAC;QAClB,CAAC;QAED;;;;;WAKG;QACH,KAAK,CAAC,MAAuB;YACzB,IAAI,cAAc,GAAG,IAAI,KAAK,EAAiB,CAAA;YAC/C,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,cAAc,CAAC,MAAM,EAAE,CAAC,EAAE,EAAC;gBAC3C,cAAc,CAAC,IAAI,CAAC,IAAI,KAAK,EAAU,CAAC,CAAC;gBACzC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,SAAS,EAAE,EAAE,CAAC,EAAE,EAAC;oBACvD,IAAI,IAAI,GAAkB,MAAM,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC;oBAC3D,cAAc,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC,CAAC;iBACzC;aACJ;YACD,IAAI,OAAO,GAAG,IAAI,GAAG,EAAU,CAAA;YAC/B,KAAK,IAAI,GAAG,IAAI,MAAM,CAAC,UAAU,EAAE,EAAC;gBAChC,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,CAAA;aACnB;YACD,IAAI,CAAC,GAAG,GAAG,IAAI,WAAI,CAAe,OAAO,EAAE,cAAc,EAAE,MAAM,CAAC,sBAAsB,EAAE,CAAC,CAAC;QAChG,CAAC;KAEJ;IA1CD,oCA0CC"}

package/dist/NaivePosTagger.d.ts

@@ -0,0 +1,21 @@
+import { PosTagger } from "./PosTagger";
+import { Sentence } from "nlptoolkit-corpus/dist/Sentence";
+import { PosTaggedCorpus } from "./PosTaggedCorpus";
+export declare class NaivePosTagger implements PosTagger {
+    private maxMap;
+    /**
+     * Test method for the Naive pos tagger. For each word, the method chooses the maximum a posterior tag from all
+     * possible tag list for that word.
+     *
+     * @param sentence Sentence to be tagged.
+     * @return Annotated (tagged) sentence.
+     */
+    posTag(sentence: Sentence): Sentence;
+    /**
+     * Train method for the Naive pos tagger. The algorithm gets all possible tag list. Then counts all
+     * possible tags (with its counts) for each possible word.
+     *
+     * @param corpus Training data for the tagger.
+     */
+    train(corpus: PosTaggedCorpus): void;
+}

package/dist/NaivePosTagger.js

@@ -0,0 +1,61 @@
+(function (factory) {
+    if (typeof module === "object" && typeof module.exports === "object") {
+        var v = factory(require, exports);
+        if (v !== undefined) module.exports = v;
+    }
+    else if (typeof define === "function" && define.amd) {
+        define(["require", "exports", "nlptoolkit-corpus/dist/Sentence", "./PosTaggedWord", "nlptoolkit-datastructure/dist/CounterHashMap"], factory);
+    }
+})(function (require, exports) {
+    "use strict";
+    Object.defineProperty(exports, "__esModule", { value: true });
+    exports.NaivePosTagger = void 0;
+    const Sentence_1 = require("nlptoolkit-corpus/dist/Sentence");
+    const PosTaggedWord_1 = require("./PosTaggedWord");
+    const CounterHashMap_1 = require("nlptoolkit-datastructure/dist/CounterHashMap");
+    class NaivePosTagger {
+        /**
+         * Test method for the Naive pos tagger. For each word, the method chooses the maximum a posterior tag from all
+         * possible tag list for that word.
+         *
+         * @param sentence Sentence to be tagged.
+         * @return Annotated (tagged) sentence.
+         */
+        posTag(sentence) {
+            let result = new Sentence_1.Sentence();
+            for (let i = 0; i < sentence.wordCount(); i++) {
+                result.addWord(new PosTaggedWord_1.PosTaggedWord(sentence.getWord(i).getName(), this.maxMap.get(sentence.getWord(i).getName())));
+            }
+            return result;
+        }
+        /**
+         * Train method for the Naive pos tagger. The algorithm gets all possible tag list. Then counts all
+         * possible tags (with its counts) for each possible word.
+         *
+         * @param corpus Training data for the tagger.
+         */
+        train(corpus) {
+            let map = new Map();
+            for (let i = 0; i < corpus.sentenceCount(); i++) {
+                let s = corpus.getSentence(i);
+                for (let j = 0; j < s.wordCount(); j++) {
+                    let word = corpus.getSentence(i).getWord(j);
+                    if (map.has(word.getName())) {
+                        map.get(word.getName()).put(word.getTag());
+                    }
+                    else {
+                        let counterMap = new CounterHashMap_1.CounterHashMap();
+                        counterMap.put(word.getTag());
+                        map.set(word.getName(), counterMap);
+                    }
+                }
+            }
+            this.maxMap = new Map();
+            for (let word of map.keys()) {
+                this.maxMap.set(word, map.get(word).max());
+            }
+        }
+    }
+    exports.NaivePosTagger = NaivePosTagger;
+});
+//# sourceMappingURL=NaivePosTagger.js.map

package/dist/NaivePosTagger.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"NaivePosTagger.js","sourceRoot":"","sources":["../source/NaivePosTagger.ts"],"names":[],"mappings":";;;;;;;;;;;;IACA,8DAAyD;IAEzD,mDAA8C;IAC9C,iFAA4E;IAE5E,MAAa,cAAc;QAIvB;;;;;;WAMG;QACH,MAAM,CAAC,QAAkB;YACrB,IAAI,MAAM,GAAG,IAAI,mBAAQ,EAAE,CAAC;YAC5B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,SAAS,EAAE,EAAE,CAAC,EAAE,EAAC;gBAC1C,MAAM,CAAC,OAAO,CAAC,IAAI,6BAAa,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,OAAO,EAAE,EAAE,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,CAAC;aACpH;YACD,OAAO,MAAM,CAAC;QAClB,CAAC;QAED;;;;;WAKG;QACH,KAAK,CAAC,MAAuB;YACzB,IAAI,GAAG,GAAG,IAAI,GAAG,EAAkC,CAAC;YACpD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,aAAa,EAAE,EAAE,CAAC,EAAE,EAAC;gBAC5C,IAAI,CAAC,GAAG,MAAM,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC;gBAC9B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,SAAS,EAAE,EAAE,CAAC,EAAE,EAAC;oBACnC,IAAI,IAAI,GAAmB,MAAM,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC;oBAC5D,IAAI,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC,EAAC;wBACxB,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC,CAAC,GAAG,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC,CAAC;qBAC9C;yBAAM;wBACH,IAAI,UAAU,GAAG,IAAI,+BAAc,EAAU,CAAC;wBAC9C,UAAU,CAAC,GAAG,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC,CAAC;wBAC9B,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,EAAE,EAAE,UAAU,CAAC,CAAC;qBACvC;iBACJ;aACJ;YACD,IAAI,CAAC,MAAM,GAAG,IAAI,GAAG,EAAkB,CAAC;YACxC,KAAK,IAAI,IAAI,IAAI,GAAG,CAAC,IAAI,EAAE,EAAC;gBACxB,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,IAAI,EAAE,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC;aAC9C;QACL,CAAC;KAEJ;IA9CD,wCA8CC"}

package/dist/PosTaggedCorpus.d.ts

@@ -0,0 +1,15 @@
+import { Corpus } from "nlptoolkit-corpus/dist/Corpus";
+export declare class PosTaggedCorpus extends Corpus {
+    private tagList;
+    /**
+     * A constructor of {@link PosTaggedCorpus} which initializes the sentences of the corpus, the word list of
+     * the corpus, and all possible tags.
+     */
+    constructor(fileName?: string);
+    /**
+     * getTagList returns all possible tags as a set.
+     *
+     * @return Set of all possible tags.
+     */
+    getTagList(): IterableIterator<string>;
+}

package/dist/PosTaggedCorpus.js

@@ -0,0 +1,75 @@
+(function (factory) {
+    if (typeof module === "object" && typeof module.exports === "object") {
+        var v = factory(require, exports);
+        if (v !== undefined) module.exports = v;
+    }
+    else if (typeof define === "function" && define.amd) {
+        define(["require", "exports", "nlptoolkit-corpus/dist/Corpus", "nlptoolkit-datastructure/dist/CounterHashMap", "nlptoolkit-corpus/dist/Sentence", "fs", "./PosTaggedWord"], factory);
+    }
+})(function (require, exports) {
+    "use strict";
+    Object.defineProperty(exports, "__esModule", { value: true });
+    exports.PosTaggedCorpus = void 0;
+    const Corpus_1 = require("nlptoolkit-corpus/dist/Corpus");
+    const CounterHashMap_1 = require("nlptoolkit-datastructure/dist/CounterHashMap");
+    const Sentence_1 = require("nlptoolkit-corpus/dist/Sentence");
+    const fs = require("fs");
+    const PosTaggedWord_1 = require("./PosTaggedWord");
+    class PosTaggedCorpus extends Corpus_1.Corpus {
+        /**
+         * A constructor of {@link PosTaggedCorpus} which initializes the sentences of the corpus, the word list of
+         * the corpus, and all possible tags.
+         */
+        constructor(fileName) {
+            super();
+            this.tagList = new CounterHashMap_1.CounterHashMap();
+            if (fileName != undefined) {
+                let newSentence = new Sentence_1.Sentence();
+                let data = fs.readFileSync(fileName, 'utf8');
+                let lines = data.split("\n");
+                for (let line of lines) {
+                    let words = line.split(/\s/);
+                    for (let word of words) {
+                        if (word != "") {
+                            if (word.includes("/")) {
+                                let name = word.substring(0, word.lastIndexOf('/'));
+                                let tag = word.substring(word.lastIndexOf('/') + 1);
+                                let shortTag;
+                                if (tag.includes("+")) {
+                                    shortTag = tag.substring(0, tag.indexOf("+"));
+                                }
+                                else {
+                                    if (tag.includes("-")) {
+                                        shortTag = tag.substring(0, tag.indexOf("-"));
+                                    }
+                                    else {
+                                        shortTag = tag;
+                                    }
+                                }
+                                this.tagList.put(shortTag);
+                                newSentence.addWord(new PosTaggedWord_1.PosTaggedWord(name, shortTag));
+                                if (tag == ".") {
+                                    this.addSentence(newSentence);
+                                    newSentence = new Sentence_1.Sentence();
+                                }
+                            }
+                        }
+                    }
+                }
+                if (newSentence.wordCount() > 0) {
+                    this.addSentence(newSentence);
+                }
+            }
+        }
+        /**
+         * getTagList returns all possible tags as a set.
+         *
+         * @return Set of all possible tags.
+         */
+        getTagList() {
+            return this.tagList.keys();
+        }
+    }
+    exports.PosTaggedCorpus = PosTaggedCorpus;
+});
+//# sourceMappingURL=PosTaggedCorpus.js.map

package/dist/PosTaggedCorpus.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"PosTaggedCorpus.js","sourceRoot":"","sources":["../source/PosTaggedCorpus.ts"],"names":[],"mappings":";;;;;;;;;;;;IAAA,0DAAqD;IACrD,iFAA4E;IAC5E,8DAAyD;IACzD,yBAAyB;IACzB,mDAA8C;IAE9C,MAAa,eAAgB,SAAQ,eAAM;QAIvC;;;WAGG;QACH,YAAY,QAAiB;YACzB,KAAK,EAAE,CAAC;YAPJ,YAAO,GAA2B,IAAI,+BAAc,EAAU,CAAA;YAQlE,IAAI,QAAQ,IAAI,SAAS,EAAE;gBACvB,IAAI,WAAW,GAAG,IAAI,mBAAQ,EAAE,CAAC;gBACjC,IAAI,IAAI,GAAG,EAAE,CAAC,YAAY,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAA;gBAC5C,IAAI,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAA;gBAC5B,KAAK,IAAI,IAAI,IAAI,KAAK,EAAE;oBACpB,IAAI,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;oBAC7B,KAAK,IAAI,IAAI,IAAI,KAAK,EAAC;wBACnB,IAAI,IAAI,IAAI,EAAE,EAAC;4BACX,IAAI,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAC;gCACnB,IAAI,IAAI,GAAG,IAAI,CAAC,SAAS,CAAC,CAAC,EAAE,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC,CAAC;gCACpD,IAAI,GAAG,GAAG,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC;gCACpD,IAAI,QAAQ,CAAA;gCACZ,IAAI,GAAG,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAC;oCAClB,QAAQ,GAAG,GAAG,CAAC,SAAS,CAAC,CAAC,EAAE,GAAG,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC;iCACjD;qCAAM;oCACH,IAAI,GAAG,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAC;wCAClB,QAAQ,GAAG,GAAG,CAAC,SAAS,CAAC,CAAC,EAAE,GAAG,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC;qCACjD;yCAAM;wCACH,QAAQ,GAAG,GAAG,CAAC;qCAClB;iCACJ;gCACD,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC;gCAC3B,WAAW,CAAC,OAAO,CAAC,IAAI,6BAAa,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAC,CAAC;gCACvD,IAAI,GAAG,IAAI,GAAG,EAAC;oCACX,IAAI,CAAC,WAAW,CAAC,WAAW,CAAC,CAAC;oCAC9B,WAAW,GAAG,IAAI,mBAAQ,EAAE,CAAC;iCAChC;6BACJ;yBACJ;qBACJ;iBACJ;gBACD,IAAI,WAAW,CAAC,SAAS,EAAE,GAAG,CAAC,EAAC;oBAC5B,IAAI,CAAC,WAAW,CAAC,WAAW,CAAC,CAAC;iBACjC;aACJ;QACL,CAAC;QAED;;;;WAIG;QACH,UAAU;YACN,OAAO,IAAI,CAAC,OAAO,CAAC,IAAI,EAAE,CAAA;QAC9B,CAAC;KACJ;IAvDD,0CAuDC"}

package/dist/PosTaggedWord.d.ts

@@ -0,0 +1,16 @@
+import { Word } from "nlptoolkit-dictionary/dist/Dictionary/Word";
+export declare class PosTaggedWord extends Word {
+    private tag;
+    /**
+     * A constructor of {@link PosTaggedWord} which takes name and tag as input and sets the corresponding attributes
+     * @param name Name of the word
+     * @param tag Tag of the word
+     */
+    constructor(name: string, tag: string);
+    /**
+     * Accessor method for tag attribute.
+     *
+     * @return Tag of the word.
+     */
+    getTag(): string;
+}

package/dist/PosTaggedWord.js

@@ -0,0 +1,35 @@
+(function (factory) {
+    if (typeof module === "object" && typeof module.exports === "object") {
+        var v = factory(require, exports);
+        if (v !== undefined) module.exports = v;
+    }
+    else if (typeof define === "function" && define.amd) {
+        define(["require", "exports", "nlptoolkit-dictionary/dist/Dictionary/Word"], factory);
+    }
+})(function (require, exports) {
+    "use strict";
+    Object.defineProperty(exports, "__esModule", { value: true });
+    exports.PosTaggedWord = void 0;
+    const Word_1 = require("nlptoolkit-dictionary/dist/Dictionary/Word");
+    class PosTaggedWord extends Word_1.Word {
+        /**
+         * A constructor of {@link PosTaggedWord} which takes name and tag as input and sets the corresponding attributes
+         * @param name Name of the word
+         * @param tag Tag of the word
+         */
+        constructor(name, tag) {
+            super(name);
+            this.tag = tag;
+        }
+        /**
+         * Accessor method for tag attribute.
+         *
+         * @return Tag of the word.
+         */
+        getTag() {
+            return this.tag;
+        }
+    }
+    exports.PosTaggedWord = PosTaggedWord;
+});
+//# sourceMappingURL=PosTaggedWord.js.map

package/dist/PosTaggedWord.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"PosTaggedWord.js","sourceRoot":"","sources":["../source/PosTaggedWord.ts"],"names":[],"mappings":";;;;;;;;;;;;IAAA,qEAAgE;IAEhE,MAAa,aAAc,SAAQ,WAAI;QAInC;;;;WAIG;QACH,YAAY,IAAY,EAAE,GAAW;YACjC,KAAK,CAAC,IAAI,CAAC,CAAC;YACZ,IAAI,CAAC,GAAG,GAAG,GAAG,CAAA;QAClB,CAAC;QAED;;;;WAIG;QACH,MAAM;YACF,OAAO,IAAI,CAAC,GAAG,CAAA;QACnB,CAAC;KACJ;IAtBD,sCAsBC"}

package/dist/PosTagger.d.ts

@@ -0,0 +1,6 @@
+import { Sentence } from "nlptoolkit-corpus/dist/Sentence";
+import { PosTaggedCorpus } from "./PosTaggedCorpus";
+export interface PosTagger {
+    train(corpus: PosTaggedCorpus): void;
+    posTag(sentence: Sentence): Sentence;
+}

package/dist/PosTagger.js

@@ -0,0 +1,13 @@
+(function (factory) {
+    if (typeof module === "object" && typeof module.exports === "object") {
+        var v = factory(require, exports);
+        if (v !== undefined) module.exports = v;
+    }
+    else if (typeof define === "function" && define.amd) {
+        define(["require", "exports"], factory);
+    }
+})(function (require, exports) {
+    "use strict";
+    Object.defineProperty(exports, "__esModule", { value: true });
+});
+//# sourceMappingURL=PosTagger.js.map

package/dist/PosTagger.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"PosTagger.js","sourceRoot":"","sources":["../source/PosTagger.ts"],"names":[],"mappings":""}

package/index.js

@@ -0,0 +1,6 @@
+export * from "./dist/DummyPosTagger"
+export * from "./dist/HmmPosTagger"
+export * from "./dist/NaivePosTagger"
+export * from "./dist/PosTaggedCorpus"
+export * from "./dist/PosTaggedWord"
+export * from "./dist/PosTagger"

package/package.json

@@ -0,0 +1,31 @@
+{
+  "name": "nlptoolkit-postagger",
+  "version": "1.0.0",
+  "description": "",
+  "main": "index.js",
+  "types": "index.js",
+  "scripts": {
+    "test": "Mocha"
+  },
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/StarlangSoftware/EnglishPosTagger-Js.git"
+  },
+  "author": "Olcay Taner Yıldız",
+  "license": "ISC",
+  "bugs": {
+    "url": "https://github.com/StarlangSoftware/EnglishPosTagger-Js/issues"
+  },
+  "homepage": "https://github.com/StarlangSoftware/EnglishPosTagger-Js#readme",
+  "devDependencies": {
+    "@types/mocha": "^9.0.0",
+    "mocha": "^9.1.3",
+    "nlptoolkit-corpus": "^1.0.0",
+    "nlptoolkit-datastructure": "^1.0.0",
+    "nlptoolkit-dictionary": "^1.0.1",
+    "nlptoolkit-hmm": "^1.0.0",
+    "nlptoolkit-math": "^1.0.0",
+    "ts-node": "^10.4.0",
+    "typescript": "^4.5.2"
+  }
+}

package/source/DummyPosTagger.ts

@@ -0,0 +1,38 @@
+import {PosTagger} from "./PosTagger";
+import {Sentence} from "nlptoolkit-corpus/dist/Sentence";
+import {PosTaggedCorpus} from "./PosTaggedCorpus";
+import {PosTaggedWord} from "./PosTaggedWord";
+
+export class DummyPosTagger implements PosTagger{
+
+    private tagList: Array<string> = new Array<string>()
+
+    /**
+     * Test method for the Dummy pos tagger. For each word, the method chooses randomly a tag from all possible
+     * tag list.
+     *
+     * @param sentence Sentence to be tagged.
+     * @return Annotated (tagged) sentence.
+     */
+    posTag(sentence: Sentence): Sentence {
+        let result = new Sentence();
+        for (let i = 0; i < sentence.wordCount(); i++){
+            result.addWord(new PosTaggedWord(sentence.getWord(i).getName(),
+                this.tagList[Math.floor(Math.random() * this.tagList.length)]));
+        }
+        return result;
+    }
+
+    /**
+     * Train method for the Dummy pos tagger. The algorithm gets all possible tag list.
+     *
+     * @param corpus Training data for the tagger.
+     */
+    train(corpus: PosTaggedCorpus): void {
+        let corpusTagList = corpus.getTagList();
+        for (let tag of corpusTagList){
+            this.tagList.push(tag)
+        }
+    }
+
+}

package/source/HmmPosTagger.ts

@@ -0,0 +1,51 @@
+import {PosTagger} from "./PosTagger";
+import {Sentence} from "nlptoolkit-corpus/dist/Sentence";
+import {PosTaggedCorpus} from "./PosTaggedCorpus";
+import {Hmm} from "nlptoolkit-hmm/dist/Hmm";
+import {Word} from "nlptoolkit-dictionary/dist/Dictionary/Word";
+import {PosTaggedWord} from "./PosTaggedWord";
+import {Hmm1} from "nlptoolkit-hmm/dist/Hmm1";
+
+export class HmmPosTagger implements PosTagger{
+
+    private hmm: Hmm<string, Word>
+
+    /**
+     * Test method for the Hmm pos tagger. For each sentence, the method uses the viterbi algorithm to produce the
+     * most possible state sequence for the given sentence.
+     *
+     * @param sentence Sentence to be tagged.
+     * @return Annotated (tagged) sentence.
+     */
+    posTag(sentence: Sentence): Sentence {
+        let result = new Sentence();
+        let tagList = this.hmm.viterbi(sentence.getWords());
+        for (let i = 0; i < sentence.wordCount(); i++){
+            result.addWord(new PosTaggedWord(sentence.getWord(i).getName(), tagList[i]));
+        }
+        return result;
+    }
+
+    /**
+     * Train method for the Hmm pos tagger. The algorithm trains an Hmm from the corpus, where corpus constitutes
+     * as an observation array.
+     *
+     * @param corpus Training data for the tagger.
+     */
+    train(corpus: PosTaggedCorpus): void {
+        let emittedSymbols = new Array<Array<string>>()
+        for (let i = 0; i < emittedSymbols.length; i++){
+            emittedSymbols.push(new Array<string>());
+            for (let j = 0; j < corpus.getSentence(i).wordCount(); j++){
+                let word = <PosTaggedWord>corpus.getSentence(i).getWord(j);
+                emittedSymbols[i].push(word.getTag());
+            }
+        }
+        let tagList = new Set<string>()
+        for (let tag of corpus.getTagList()){
+            tagList.add(tag)
+        }
+        this.hmm = new Hmm1<string, Word>(tagList, emittedSymbols, corpus.getAllWordsAsArrayList());
+    }
+
+}

package/source/NaivePosTagger.ts

@@ -0,0 +1,53 @@
+import {PosTagger} from "./PosTagger";
+import {Sentence} from "nlptoolkit-corpus/dist/Sentence";
+import {PosTaggedCorpus} from "./PosTaggedCorpus";
+import {PosTaggedWord} from "./PosTaggedWord";
+import {CounterHashMap} from "nlptoolkit-datastructure/dist/CounterHashMap";
+
+export class NaivePosTagger implements PosTagger{
+
+    private maxMap: Map<string, string>
+
+    /**
+     * Test method for the Naive pos tagger. For each word, the method chooses the maximum a posterior tag from all
+     * possible tag list for that word.
+     *
+     * @param sentence Sentence to be tagged.
+     * @return Annotated (tagged) sentence.
+     */
+    posTag(sentence: Sentence): Sentence {
+        let result = new Sentence();
+        for (let i = 0; i < sentence.wordCount(); i++){
+            result.addWord(new PosTaggedWord(sentence.getWord(i).getName(), this.maxMap.get(sentence.getWord(i).getName())));
+        }
+        return result;
+    }
+
+    /**
+     * Train method for the Naive pos tagger. The algorithm gets all possible tag list. Then counts all
+     * possible tags (with its counts) for each possible word.
+     *
+     * @param corpus Training data for the tagger.
+     */
+    train(corpus: PosTaggedCorpus): void {
+        let map = new Map<string, CounterHashMap<string>>();
+        for (let i = 0; i < corpus.sentenceCount(); i++){
+            let s = corpus.getSentence(i);
+            for (let j = 0; j < s.wordCount(); j++){
+                let word = <PosTaggedWord> corpus.getSentence(i).getWord(j);
+                if (map.has(word.getName())){
+                    map.get(word.getName()).put(word.getTag());
+                } else {
+                    let counterMap = new CounterHashMap<string>();
+                    counterMap.put(word.getTag());
+                    map.set(word.getName(), counterMap);
+                }
+            }
+        }
+        this.maxMap = new Map<string, string>();
+        for (let word of map.keys()){
+            this.maxMap.set(word, map.get(word).max());
+        }
+    }
+
+}