HW2.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "Objective:\n",
        "The objective of this homework assignment is to deepen your understanding of the Word2Vec algorithm, its implementation, and its applications in natural language processing.\n",
        "\n",
        "#Instructions:\n",
        "\n",
        "* Implement Word2Vec using the gensim library in Python. You can choose a preprocessed text dataset or use your own.\n",
        "* Train Word2Vec on the dataset and experiment with different hyperparameters such as window size, vector size, and training epochs.\n",
        "* Visualize the learned word embeddings using techniques like t-SNE (t-distributed Stochastic Neighbor Embedding) to visualize high-dimensional data in a lower-dimensional space.\n",
        "\n",
        "\n",
        "\n",
        "#Application Section:\n",
        "\n",
        "* Choose one or more applications of Word2Vec (e.g., sentiment analysis, document similarity, recommendation systems) and apply the learned word embeddings to solve a relevant problem. You need to find your own data for this task.\n",
        "\n",
        "* Evaluate the performance of your Word2Vec model on the chosen application(s) and discuss the results.\n",
        "\n",
        "# Discussion Section:\n",
        "\n",
        "Reflect on the strengths and weaknesses of Word2Vec. Discuss any limitations or challenges you encountered during the implementation and application of Word2Vec.\n",
        "\n",
        "\n",
        "#Submission Guidelines:\n",
        "\n",
        "* Submit your Jupyter Notebook (.ipynb) containing the implementation, results, and discussions.\n",
        "\n",
        "* Include necessary comments and explanations in your code to ensure clarity.\n",
        "Provide clear visualizations and interpretations of your results.\n",
        "\n",
        "* The submission should be individual work. Any collaboration or external resources used should be properly cited.\n",
        "\n",
        "\n",
        "#Grading Criteria:\n",
        "Correct implementation of Word2Vec using gensim: 50%\n",
        "Effective application of Word2Vec to solve a relevant problem: 30%\n",
        "Thoughtful discussion on Word2Vec's strengths, weaknesses, and potential improvements: 20%\n",
        "\n"
      ],
      "metadata": {
        "id": "PyXptH14LjHa"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "FAeSmBJEK-P4",
        "outputId": "2edfe6d8-4909-441e-ff48-064b043a333a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:gensim.models.word2vec:Effective 'alpha' higher than previous training cycles\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Embedding for 'example': [ 9.45401989e-05  3.07740690e-03 -6.81256410e-03 -1.37506088e-03\n",
            "  7.66831636e-03  7.34624360e-03 -3.67307314e-03  2.64320313e-03\n",
            " -8.31757765e-03  6.20545540e-03 -4.63707699e-03 -3.16425879e-03\n",
            "  9.31136403e-03  8.73725978e-04  7.49065029e-03 -6.07430004e-03\n",
            "  5.16077364e-03  9.92311444e-03 -8.45800247e-03 -5.13627706e-03\n",
            " -7.06465496e-03 -4.86252457e-03 -3.77832795e-03 -8.53637140e-03\n",
            "  7.95610994e-03 -4.84422268e-03  8.42377450e-03  5.26298024e-03\n",
            " -6.55055279e-03  3.95777682e-03  5.46991872e-03 -7.42676388e-03\n",
            " -7.40543753e-03 -2.47564749e-03 -8.62605032e-03 -1.58170541e-03\n",
            " -4.03049809e-04  3.29947704e-03  1.44191808e-03 -8.81674408e-04\n",
            " -5.59466239e-03  1.73065183e-03 -8.97819933e-04  6.79364381e-03\n",
            "  3.97368614e-03  4.52956837e-03  1.43396843e-03 -2.69945408e-03\n",
            " -4.36667772e-03 -1.03165093e-03  1.43666589e-03 -2.64585507e-03\n",
            " -7.07415771e-03 -7.80545827e-03 -9.12159774e-03 -5.93553204e-03\n",
            " -1.84725015e-03 -4.32430767e-03 -6.46100147e-03 -3.71695636e-03\n",
            "  4.28918703e-03 -3.73911951e-03  8.37817974e-03  1.53365743e-03\n",
            " -7.24257063e-03  9.43415053e-03  7.63135683e-03  5.49368607e-03\n",
            " -6.84914738e-03  5.82293188e-03  4.00945684e-03  5.18590258e-03\n",
            "  4.25593508e-03  1.93935412e-03 -3.17002763e-03  8.35407805e-03\n",
            "  9.61277820e-03  3.79265659e-03 -2.83719297e-03  6.70191002e-06\n",
            "  1.21880672e-03 -8.45839456e-03 -8.22388474e-03 -2.30678284e-04\n",
            "  1.23704050e-03 -5.74343139e-03 -4.72509349e-03 -7.34648807e-03\n",
            "  8.32866877e-03  1.21631485e-04 -4.50942991e-03  5.70185436e-03\n",
            "  9.18040890e-03 -4.09996184e-03  7.96534494e-03  5.37581230e-03\n",
            "  5.87883359e-03  5.12260420e-04  8.21333006e-03 -7.01890187e-03]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1600x1600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Similarity between 'word1' and 'word2': -0.052298486\n"
          ]
        }
      ],
      "source": [
        "# use the following base code to help you in the homework\n",
        "\n",
        "# Import necessary libraries\n",
        "import gensim\n",
        "from gensim.models import Word2Vec\n",
        "from sklearn.manifold import TSNE\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "# Load your dataset (or use a preprocessed one)\n",
        "# Example: sentences = [['this', 'is', 'an', 'example', 'sentence'], ['another', 'example', 'sentence']]\n",
        "# Replace this with your own dataset\n",
        "sentences = [['this', 'is', 'an', 'example', 'sentence'], ['another', 'example', 'sentence'],  ['yet', 'another', 'sentence']]\n",
        "\n",
        "# Train Word2Vec model\n",
        "model = Word2Vec(sentences, vector_size=100, window=5, min_count=1, workers=4)\n",
        "model.train(sentences, total_examples=len(sentences), epochs=10)\n",
        "\n",
        "# Get word embeddings\n",
        "word_embeddings = model.wv\n",
        "\n",
        "# Example of accessing word embeddings\n",
        "print(\"Embedding for 'example':\", word_embeddings['example'])\n",
        "\n",
        "# Visualize word embeddings using t-SNE\n",
        "def tsne_plot(model):\n",
        "    labels = []\n",
        "    wordvecs = []\n",
        "\n",
        "    for word in model.wv.key_to_index:\n",
        "        wordvecs.append(model.wv[word])\n",
        "        labels.append(word)\n",
        "\n",
        "    tsne_model = TSNE(perplexity=2, n_components=2, init='pca', n_iter=2500, random_state=23)\n",
        "    new_values = tsne_model.fit_transform(np.array(wordvecs))\n",
        "\n",
        "    x = []\n",
        "    y = []\n",
        "    for value in new_values:\n",
        "        x.append(value[0])\n",
        "        y.append(value[1])\n",
        "\n",
        "    plt.figure(figsize=(16, 16))\n",
        "    for i in range(len(x)):\n",
        "        plt.scatter(x[i], y[i])\n",
        "        plt.annotate(labels[i], xy=(x[i], y[i]), xytext=(5, 2), textcoords='offset points', ha='right', va='bottom')\n",
        "    plt.show()\n",
        "\n",
        "# Plot t-SNE visualization\n",
        "tsne_plot(model)\n",
        "\n",
        "# Example application: similarity between words\n",
        "similarity = model.wv.similarity('example', 'sentence')\n",
        "print(\"Similarity between 'word1' and 'word2':\", similarity)\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "T-jv6QRLK_8y"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}