{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1638b7b97e3bd6f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-11-22T11:40:21.711998Z",
     "start_time": "2025-11-22T11:40:20.129376Z"
    }
   },
   "outputs": [],
   "source": [
    "import spacy\n",
    "nlp =  spacy.load(\"en_core_web_lg\") # Medium model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b79941bf4553fd6",
   "metadata": {},
   "source": [
    "Test word vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "8a3c4314a90086fe",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-11-22T11:47:39.286432Z",
     "start_time": "2025-11-22T11:47:39.271377Z"
    }
   },
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "[E010] Word vectors set to length 0. This may be because you don't have a model installed or loaded, or because your model doesn't include word vectors. For more info, see the docs:\nhttps://spacy.io/usage/models",
     "output_type": "error",
     "traceback": [
      "\u001b[31m---------------------------------------------------------------------------\u001b[39m",
      "\u001b[31mValueError\u001b[39m                                Traceback (most recent call last)",
      "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[9]\u001b[39m\u001b[32m, line 9\u001b[39m\n\u001b[32m      7\u001b[39m \u001b[38;5;28;01mfor\u001b[39;00m word2 \u001b[38;5;129;01min\u001b[39;00m words:\n\u001b[32m      8\u001b[39m     \u001b[38;5;28;01mif\u001b[39;00m word1 != word2:\n\u001b[32m----> \u001b[39m\u001b[32m9\u001b[39m         similarity = \u001b[43mnlp\u001b[49m\u001b[43m.\u001b[49m\u001b[43mvocab\u001b[49m\u001b[43m[\u001b[49m\u001b[43mword1\u001b[49m\u001b[43m]\u001b[49m\u001b[43m.\u001b[49m\u001b[43msimilarity\u001b[49m\u001b[43m(\u001b[49m\u001b[43mnlp\u001b[49m\u001b[43m.\u001b[49m\u001b[43mvocab\u001b[49m\u001b[43m[\u001b[49m\u001b[43mword2\u001b[49m\u001b[43m]\u001b[49m\u001b[43m)\u001b[49m\n\u001b[32m     10\u001b[39m         \u001b[38;5;28mprint\u001b[39m(\u001b[33mf\u001b[39m\u001b[33m\"\u001b[39m\u001b[38;5;132;01m{\u001b[39;00mword1\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m - \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mword2\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00msimilarity\u001b[38;5;132;01m:\u001b[39;00m\u001b[33m.3f\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[33m\"\u001b[39m)\n",
      "\u001b[36mFile \u001b[39m\u001b[32m~/code/plagiarism-detector/.venv/lib/python3.13/site-packages/spacy/lexeme.pyx:146\u001b[39m, in \u001b[36mspacy.lexeme.Lexeme.similarity\u001b[39m\u001b[34m()\u001b[39m\n",
      "\u001b[36mFile \u001b[39m\u001b[32m~/code/plagiarism-detector/.venv/lib/python3.13/site-packages/spacy/lexeme.pyx:164\u001b[39m, in \u001b[36mspacy.lexeme.Lexeme.vector_norm.__get__\u001b[39m\u001b[34m()\u001b[39m\n",
      "\u001b[36mFile \u001b[39m\u001b[32m~/code/plagiarism-detector/.venv/lib/python3.13/site-packages/spacy/lexeme.pyx:176\u001b[39m, in \u001b[36mspacy.lexeme.Lexeme.vector.__get__\u001b[39m\u001b[34m()\u001b[39m\n",
      "\u001b[31mValueError\u001b[39m: [E010] Word vectors set to length 0. This may be because you don't have a model installed or loaded, or because your model doesn't include word vectors. For more info, see the docs:\nhttps://spacy.io/usage/models"
     ]
    }
   ],
   "source": [
    "def test_word_vectors(word):\n",
    "    print(word, nlp.vocab[word].vector.shape)\n",
    "\n",
    "words = [\"cat\", \"dog\", \"feline\", \"feral\", \"vehicle\", \"car\"]\n",
    "# Test work similarities\n",
    "for word1 in words:\n",
    "    for word2 in words:\n",
    "        if word1 != word2:\n",
    "            similarity = nlp.vocab[word1].similarity(nlp.vocab[word2])\n",
    "            print(f\"{word1} - {word2}: {similarity:.3f}\")\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8f32b5695f554268",
   "metadata": {},
   "source": [
    "Simple averaging"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "68a6757447e4a1c7",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-11-18T23:45:03.085563Z",
     "start_time": "2025-11-18T23:45:03.082190Z"
    }
   },
   "outputs": [],
   "source": [
    "def sentence_similarity_avg(sent1, sent2):\n",
    "    doc1 = nlp(sent1)\n",
    "    doc2 = nlp(sent2)\n",
    "\n",
    "    # Vectors for each word, filter out words without vectors (medium model)\n",
    "    vecs1 = [token.vector for token in doc1 if token.has_vector]\n",
    "    vecs2 = [token.vector for token in doc2 if token.has_vector]\n",
    "\n",
    "    if not vecs1 or not vecs2:\n",
    "        return 0.0\n",
    "\n",
    "    # Average vectors\n",
    "    avg1 = sum(vecs1) / len(vecs1)\n",
    "    avg2 = sum(vecs2) / len(vecs2)\n",
    "\n",
    "    #cosine similarity\n",
    "    from sklearn.metrics.pairwise import cosine_similarity\n",
    "    return cosine_similarity([avg1], [avg2])[0][0]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9c3aa050f5bc0fe",
   "metadata": {},
   "source": [
    "SIF - Smooth Inverse Similarity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c100956f89d9b581",
   "metadata": {},
   "outputs": [],
   "source": [
    "def sentence_similarity_sif(sent1, sent2):\n",
    "    doc1 = nlp(sent1)\n",
    "    doc2 = nlp(sent2)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.13.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}