Week_1_NLP_Fundamentals_&_Word_Embeddings.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": [
        "# Welcome to NLP in 3 weeks\n",
        "\n",
        "This course is about the power and diversity of NLP methods. We will start with what NLP was in the past and progress to state of the art Transformer models. Through this journey we will discuss how and why methods work!"
      ],
      "metadata": {
        "id": "-csp4xSIoOyU"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![nlp_in_3_weeks.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "lv8A21UmpnH5"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "We will also address the following important questions:\n",
        "\n",
        "\n",
        "1.   What is an Embedding (as in Word Embedding or Sentence Embedding)?\n",
        "2.   What is a Transformer Model?\n",
        "3.   How can I use Transformer models to solve modern day NLP problems?\n",
        "4.   When do I even need to train my own model - given these great OpenAI (etc.) offerings?\n",
        "\n"
      ],
      "metadata": {
        "id": "UuLFmyYCt-PQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Setup\n",
        "# This cell installs all the dependencies used in this week - imports happen as we go\n",
        "%%capture\n",
        "import nltk\n",
        "nltk.download('punkt')\n",
        "nltk.download('stopwords')\n",
        "nltk.download('averaged_perceptron_tagger')\n",
        "nltk.download('wordnet')\n",
        "nltk.download('omw-1.4')\n",
        "nltk.download('word2vec_sample')\n",
        "!pip install datasets"
      ],
      "metadata": {
        "id": "p6Y8kmX3GdPU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Defining NLP"
      ],
      "metadata": {
        "id": "o9kzzqLh1r24"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## What is NLP?\n",
        "\n",
        "*Natural language processing (NLP) is a subfield of linguistics, computer science, and artificial intelligence concerned with the interactions between computers and human language, in particular how to program computers to process and analyze large amounts of natural language data. The goal is a computer capable of **understanding** the contents of documents, including the contextual nuances of the language within them. The technology can then accurately extract information and insights contained in the documents as well as categorize and organize the documents themselves.*\n",
        "\n",
        "A great summary from [Wikipedia](https://en.wikipedia.org/w/index.php?title=Natural_language_processing&oldid=1122895538)"
      ],
      "metadata": {
        "id": "PZVQnYy2nC1X"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## What is a natural language?\n",
        "\n",
        "It is a language that has been evolving naturally as a mean of communicaton between humans.\n",
        "\n",
        "\n",
        "---\n",
        "\n",
        "### A few problems to ponder about\n",
        "\n",
        "*   Every human speaks a little bit different\n",
        "*   Sometimes we have missunderstandings - even if we use the exact same words, we might describe something different\n",
        "*   Our Vocabulary evolves over time\n",
        "*   Sometimes we mean the same using, but are using different words, expressions and or even languages\n",
        "*   Per above's defintion: is ChatGPT NLP?"
      ],
      "metadata": {
        "id": "Lym-jcYZnWfX"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## How we thought we can teach computers language"
      ],
      "metadata": {
        "id": "ayBR11U3oiqE"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Symbolic NLP\n",
        "\n",
        "Formulize all languages as a set of rules (Grammar) and try to create a computational graph from these rules and relevant objects (words / entities).\n",
        "\n",
        "\n",
        "\n",
        "*   ❌ It is very hard to come up with rules for every person and situtation.\n",
        "*   ❌ Changes to langugae mean changes to the rules.\n"
      ],
      "metadata": {
        "id": "fbX5eROoow38"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Statistical NLP\n",
        "\n",
        "Language is conditioned by its wider context (audience / corpus). Instead of fixed rules we try to extract rules from statistical distributions and relationships in that limited context.\n",
        "\n",
        "*   ✅ Changes over time do not have to be incorporated manually\n",
        "*   ❌ Deciding which dsitributions are of importance remains a manual task\n",
        "*   ❌ Still has a hard time conceptualizing dual meanings of words and or expressions"
      ],
      "metadata": {
        "id": "4J9BRvR0pPjX"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Neural NLP\n",
        "\n",
        "All features of a language (grammar, words, semantics) are extracted from the used corpus in an automatic fashion. High-paramterized neural networks can operate on a semantic representation of language to yield the desired output.\n",
        "\n",
        "\n",
        "*   This is where we are now and if there are any drawbacks, they are...\n",
        "*   ... that we need a lot of ressources for these neural network models.\n",
        "\n"
      ],
      "metadata": {
        "id": "-y2Y1F4erPbE"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Teaching a computer natural language\n",
        "\n",
        "Luckily, representing text is already a solved problem. In most programming languages there is a powerful \"String\" implementation and Unicode standardises even the representation of Emoji 🦄!\n",
        "\n",
        "[Every Emoji by Codepoint](https://emojipedia.org/emoji/)"
      ],
      "metadata": {
        "id": "QeGUI6jr1pvN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print('\\U0001f4af')"
      ],
      "metadata": {
        "id": "FJEC6Urcthjf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print('\\u2764\\ufe0f')"
      ],
      "metadata": {
        "id": "wQ8ZFibPvaLf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "But characters are probably not what language is actually about - at least in English!"
      ],
      "metadata": {
        "id": "wEAGnYslwUob"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## The  Twitter Sentiment Analysis Dataset\n",
        "\n",
        "To load individual words with semantic meaning or any value at all we need a datastructure that allows us to see words as a unite compared to an array of characters...\n",
        "\n",
        "\n",
        "In the following section we want to use methods from symbolic and statistical NLP to make predictions on texts. To be more precise we want to take a look at the [Twitter Sentiment Analysis](https://huggingface.co/datasets/carblacac/twitter-sentiment-analysis) collected by University of Michigan and a Kaggle Twitter Sentiment Corpus by Niek Sanders. This dataset is a scrape of Twitter and tweets were labeled as positive (1) or  negative (0).\n",
        "\n",
        "For this mini project we will make use of the [nltk Python library](https://www.nltk.org) that implements many ideas from symbolic and statistical NLP.\n",
        "\n",
        "There is also the [spaCy Python library](https://spacy.io/) which has even more functionality, but tends to be slower."
      ],
      "metadata": {
        "id": "BQ0Y51U__WNp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's import nltk and pandas\n",
        "# We actually already imported nltk before in the setup cell and downloaded a bunch of\n",
        "# dependencies.\n",
        "\n",
        "import pandas as pd\n",
        "import nltk"
      ],
      "metadata": {
        "id": "mQUCUze1CojL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Download the dataset and create a Pandas Dataframe\n",
        "%%capture\n",
        "from datasets import load_dataset\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "\n",
        "dataset = load_dataset(\"carblacac/twitter-sentiment-analysis\", \"None\")\n",
        "# Stratified Sample of Dataset\n",
        "datapoints = 10000\n",
        "\n",
        "X_train, _, y_train, _ = train_test_split(dataset[\"validation\"][\"text\"], dataset[\"validation\"][\"feeling\"],\n",
        "                                                    train_size=datapoints, random_state=0,\n",
        "                                                    stratify=dataset[\"validation\"][\"feeling\"])\n",
        "\n",
        "df = pd.DataFrame({\"text\": X_train, \"positive\": y_train})\n",
        "df[\"positive\"] = df[\"positive\"].astype(\"bool\")\n",
        "\n",
        "# We save some memory by deallocating the downloaded dataset\n",
        "dataset = None"
      ],
      "metadata": {
        "id": "fqFdCQGzDLk5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's peek into the dataset\n",
        "df"
      ],
      "metadata": {
        "id": "9IE0y6KQCzvw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Extract a single observation\n",
        "sample =  df.sample(1, random_state=34)\n",
        "text = sample[\"text\"].iat[0].split(\".\")\n",
        "positive = sample[\"positive\"].iat[0]\n",
        "\n",
        "\n",
        "for line in text:\n",
        "    print(line)\n",
        "\n",
        "print(f\"Positive Feeling? - {positive}\")"
      ],
      "metadata": {
        "id": "jvTiv8DpD2In"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Symbolic NLP - Explorartory Data Analysis\n",
        "Just how complicated are words?\n",
        "\n",
        "First let's make a list of all the words!"
      ],
      "metadata": {
        "id": "REGIOwxj2Qzo"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# First intuition: Let's count word occourences\n",
        "from collections import defaultdict, Counter\n",
        "from tqdm import tqdm\n",
        "\n",
        "# We use Python's build in Counter object here\n",
        "word_counts = Counter()\n",
        "\n",
        "for row in tqdm(df.iterrows(), total=len(df)):\n",
        "    text = row[1]\n",
        "\n",
        "    # Pandas's built in split() method of a string column splits on whitespace characters\n",
        "    for word in text.str.split()[0]:\n",
        "        word_counts[word] += 1\n",
        "\n",
        "len(word_counts)"
      ],
      "metadata": {
        "id": "geHiWcgo0Y0G"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Peek at the most common 25 words\n",
        "word_counts.most_common(25)"
      ],
      "metadata": {
        "id": "KxlS93tsE4mw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Peek at 10 of the words that only appeared once\n",
        "# Note the words ending in fullstops or words connected by punctuation.\n",
        "word_counts.most_common()[:-len(word_counts)-1:-1][10:20]"
      ],
      "metadata": {
        "id": "f5FUjgwIFW10"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Remove punctuation"
      ],
      "metadata": {
        "id": "iEfowhknwclM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# NLTK has a split function that uses rules - e.g. implements symbolic NLP\n",
        "from nltk import word_tokenize"
      ],
      "metadata": {
        "id": "nWDOozq-GP3K"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts = Counter()\n",
        "\n",
        "for row in tqdm(df.iterrows(), total=len(df)):\n",
        "    text = row[1].iat[0]\n",
        "    for word in word_tokenize(text):\n",
        "        word_counts[word] += 1\n",
        "\n",
        "len(word_counts)"
      ],
      "metadata": {
        "id": "ZvSNP5-YGTC8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Using NLTK's word_tokenize we could get rid of most punctuation\n",
        "word_counts.most_common()[:-len(word_counts)-1:-1][10:20]"
      ],
      "metadata": {
        "id": "1aeBYPkTG07q"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Make things lowercase"
      ],
      "metadata": {
        "id": "zd7BNTtUwk9Z"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "At this point of practicing symbolic NLP you might get an idea how much manual labor is necessary to come up with all the rules needed.\n",
        "\n",
        "One thing we notice with the English language is that most words are not cased. E.g. upper and lower case has hardly any meaning and instead is more of a decorative feature.\n",
        "\n",
        "What about your language?\n",
        "\n",
        "In German the meaning of lowercase and uppercase words is more important when words are isolated. In Chinese there is no whitespace nor upper and lower case at all..."
      ],
      "metadata": {
        "id": "vge-TQbYcaDQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's turn everything lowercase\n",
        "word_counts = Counter()\n",
        "\n",
        "for row in tqdm(df.iterrows(), total=len(df)):\n",
        "    text = row[1].iat[0]\n",
        "\n",
        "    for word in word_tokenize(text):\n",
        "        word_lowercase = word.lower()\n",
        "        word_counts[word_lowercase] += 1\n",
        "\n",
        "len(word_counts)"
      ],
      "metadata": {
        "id": "QycLxb7LG-jI"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts.most_common()[:-len(word_counts)-1:-1][10:20]"
      ],
      "metadata": {
        "id": "6mEk4bSXHJ46"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### A hint of statistical NLP - remove Stopwords and special characters\n",
        "The most common words are probably less useful for predicting whether somebody wrote a positive or negative tweet - luckily, there are handy tools to remove them provided by NLTK."
      ],
      "metadata": {
        "id": "sRnE6ff3dkns"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from nltk.corpus import stopwords\n",
        "\n",
        "stopwords_en = stopwords.words('english')\n",
        "\n",
        "word_counts = Counter()\n",
        "\n",
        "for row in tqdm(df.iterrows(), total=len(df)):\n",
        "    text = row[1].iat[0]\n",
        "\n",
        "    for word in word_tokenize(text):\n",
        "        word_lowercase = word.lower()\n",
        "\n",
        "        if word_lowercase not in stopwords_en:\n",
        "            word_counts[word_lowercase] += 1\n",
        "\n",
        "len(word_counts)"
      ],
      "metadata": {
        "id": "PFgRGDVDHO79"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts.most_common()[:-len(word_counts)-1:-1][10:20]"
      ],
      "metadata": {
        "id": "wrUXcdFQHydc"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts.most_common(10)"
      ],
      "metadata": {
        "id": "8s5S9TnoH3dI"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Oh no, hidden among our most common words are isolated special characters..."
      ],
      "metadata": {
        "id": "J2XBaCY_d585"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from string import punctuation\n",
        "\n",
        "# We can add Python's native list of punctuation characters to our stopwords using a set union\n",
        "stopwords_en = set(stopwords_en).union(set(punctuation))\n",
        "\n",
        "word_counts = Counter()\n",
        "\n",
        "for row in tqdm(df.iterrows(), total=len(df)):\n",
        "    text = row[1].iat[0]\n",
        "\n",
        "    for word in word_tokenize(text):\n",
        "        word_lowercase = word.lower()\n",
        "\n",
        "        if word_lowercase not in stopwords_en:\n",
        "            word_counts[word_lowercase] += 1\n",
        "\n",
        "len(word_counts)"
      ],
      "metadata": {
        "id": "rLbXtvwVH8XU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts.most_common(10)"
      ],
      "metadata": {
        "id": "ph39poeBIW1_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Stemming and Lemmatization\n",
        "\n",
        "There are two more operations we could do to \"simplify\" our vocabulary...\n",
        "\n",
        "\n",
        "\n",
        "*   **Stemming**: Remove all suffixes, declianations and conjugations from a word and only leave the stem\n",
        "    *   meeting --> meet\n",
        "    *   meets --> meet\n",
        "    *   meet --> meet\n",
        "\n",
        "*   **Lemmatization**: First identify the part of speech of a word, than remove inflictions\n",
        "    * meeting (Verb) --> meet\n",
        "    * meeting (Noun) --> Meeting\n",
        "\n",
        "How do we know the the part of speech of a word? Symbolic NLP - just like in school 🙈\n"
      ],
      "metadata": {
        "id": "tfWDqqc-eXj6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# This cell uses Stemming\n",
        "from nltk.stem import PorterStemmer\n",
        "\n",
        "porter = PorterStemmer()\n",
        "\n",
        "word_counts = Counter()\n",
        "\n",
        "for row in tqdm(df.iterrows(), total=len(df)):\n",
        "    text = row[1].iat[0]\n",
        "\n",
        "    for word in word_tokenize(text):\n",
        "        word_lowercase = word.lower()\n",
        "\n",
        "        if word_lowercase not in stopwords_en:\n",
        "            stemmed_word = porter.stem(word_lowercase)\n",
        "            word_counts[stemmed_word] += 1\n",
        "\n",
        "len(word_counts)"
      ],
      "metadata": {
        "id": "zDJGciutIh8-"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's see stemming at work in detail\n",
        "print(\"meeting --> \", porter.stem(\"meeting\"))\n",
        "print(\"meets --> \", porter.stem(\"meets\"))\n",
        "print(\"the meeting --> \", porter.stem(\"the meeting\"))"
      ],
      "metadata": {
        "id": "KSkeG3UNQf5m"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts.most_common(10)"
      ],
      "metadata": {
        "id": "yXIdBRmQI9Iz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# And this cell uses Lemmatization\n",
        "\n",
        "from nltk import pos_tag\n",
        "from nltk.stem import WordNetLemmatizer\n",
        "\n",
        "wnl = WordNetLemmatizer()\n",
        "\n",
        "def penn2morphy(penntag):\n",
        "    \"\"\" Converts Penn Treebank tags to WordNet. \"\"\"\n",
        "    morphy_tag = {'NN':'n', 'JJ':'a',\n",
        "                  'VB':'v', 'RB':'r'}\n",
        "    try:\n",
        "        return morphy_tag[penntag[:2]]\n",
        "    except:\n",
        "        return 'n'\n",
        "\n",
        "word_counts = Counter()\n",
        "\n",
        "for row in tqdm(df.iterrows(), total=len(df)):\n",
        "    text = row[1].iat[0]\n",
        "\n",
        "    for word, tag in pos_tag(word_tokenize(text)):\n",
        "        word_lemmatized = wnl.lemmatize(word.lower(), pos=penn2morphy(tag))\n",
        "\n",
        "        if word_lemmatized not in stopwords_en:\n",
        "            word_counts[word_lemmatized] += 1\n",
        "\n",
        "len(word_counts)"
      ],
      "metadata": {
        "id": "Ai7SUsvIJw9E"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "word_counts.most_common(10)"
      ],
      "metadata": {
        "id": "HByrrzpeKoeL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's see lemmatization in detail\n",
        "text = \"\"\"Deep in the forest, the council is meeting for the first time since the 'event'.\\n\n",
        "          Right when the meeting commences, a long howl disrupts and snaps them to immediate attention and the mages ...\"\"\"\n",
        "\n",
        "text_lemmatized = \"\"\n",
        "for word, tag in pos_tag(word_tokenize(text)):\n",
        "    word_lemmatized = wnl.lemmatize(word.lower(), pos=penn2morphy(tag))\n",
        "    text_lemmatized += f\"{word} ({tag}, {word_lemmatized}) \"\n",
        "\n",
        "print(text_lemmatized)"
      ],
      "metadata": {
        "id": "U_vXT2tpQlqL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "If you want to learn more about nltk's pos tagger - here is an interesting [Blog Post by the author](https://explosion.ai/blog/part-of-speech-pos-tagger-in-python)."
      ],
      "metadata": {
        "id": "LPJlhcjwWm66"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Statistical NLP - Training Machine Learning models on texts\n",
        "\n",
        "This is what you came for, right, maybe?\n",
        "\n",
        "Let us define a very simple base model - we want to model the relationship between the number of times \"love\" appeared in a text and whether the tweet expressed a positive sentiment."
      ],
      "metadata": {
        "id": "nn4d_0bbgpeb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "X = []\n",
        "y = []\n",
        "\n",
        "for row in tqdm(df.iterrows(), total=len(df)):\n",
        "    text = row[1].iat[0]\n",
        "    x = 0\n",
        "\n",
        "    for word, tag in pos_tag(word_tokenize(text)):\n",
        "        word_lemmatized = wnl.lemmatize(word.lower(), pos=penn2morphy(tag))\n",
        "\n",
        "        if word_lemmatized == \"love\":\n",
        "            x += 1\n",
        "\n",
        "    X.append(x)\n",
        "    y.append(row[1].iat[1])"
      ],
      "metadata": {
        "id": "CtwYqiVbLxcu"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's define a meaningful train-test-split\n",
        "from sklearn.model_selection import train_test_split\n",
        "import numpy as np\n",
        "\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(np.array(X).reshape(-1, 1), np.array(y),\n",
        "                                                    test_size=0.15, random_state=0,\n",
        "                                                    stratify=y)\n",
        "\n",
        "len(y_train), len(y_test)"
      ],
      "metadata": {
        "id": "zG0WcShqiSXg"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "\n",
        "clf = LogisticRegression()\n",
        "clf.fit(X=X_train, y=y_train)"
      ],
      "metadata": {
        "id": "_UuzPljFMs0P"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "clf.score(X=X_test, y=y_test) # 😑"
      ],
      "metadata": {
        "id": "cZGCjrv-NFqk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Simple Word Vectors\n",
        "\n",
        "This model is obviously useless - we threw away most of the information available!\n",
        "\n",
        "We have to consider other words, too. The simplest approach is using a bag-of-words approach. E. g. for each text we use the counts of each word, but with no respect to their position in the sentence. Just like a word cloud!\n",
        "\n",
        "Since we have to introduce one predictor per word to our model we basically treat them as categorical data. Instead of counting the words we might also just one hot encode them.\n",
        "\n",
        "The process of converting words to vectors is called word emebdding and this is the most simple version of it."
      ],
      "metadata": {
        "id": "LwUpPJK5jPQI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's first define a meaningful train-test-split\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(df[\"text\"],\n",
        "                                                    df[\"positive\"],\n",
        "                                                    test_size=0.15, random_state=0,\n",
        "                                                    stratify=df[\"positive\"])\n",
        "\n",
        "len(y_train), len(y_test)"
      ],
      "metadata": {
        "id": "SEDdrnhC-Hi7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# From EDA above we can formulate this function that extracts all the words we care about\n",
        "# from a given text\n",
        "def preprocessing_function(text):\n",
        "    words = []\n",
        "\n",
        "    for word, tag in pos_tag(word_tokenize(text)):\n",
        "        word_lemmatized = wnl.lemmatize(word.lower(), pos=penn2morphy(tag))\n",
        "\n",
        "        if word_lemmatized not in stopwords_en and not word_lemmatized.isdigit():\n",
        "            words.append(word_lemmatized)\n",
        "\n",
        "    return words"
      ],
      "metadata": {
        "id": "AIzAaTv3_2Xm"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Sklearn has a CountVectorizer object that allows us to pass an \"analyzer\"\n",
        "# The \"analyzer\" gets the text passed and returns a list of things to count\n",
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "\n",
        "count_vectorizer = CountVectorizer(analyzer=preprocessing_function)"
      ],
      "metadata": {
        "id": "RjdooQGT_b7W"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# We train the CountVectorizer on the train set and transform both datsets\n",
        "X_train_vectorized = count_vectorizer.fit_transform(X_train)\n",
        "X_test_vectorized = count_vectorizer.transform(X_test)"
      ],
      "metadata": {
        "id": "aiFfV-PTBIfj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "![count_vector.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "pvv193mtrvmQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Naturally, the resulting vector is pretty sparse\n",
        "X_train_vectorized.toarray()[0].shape"
      ],
      "metadata": {
        "id": "vvvkcgiqoF66"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "count_vectorizer.vocabulary_"
      ],
      "metadata": {
        "id": "9oO4Vuvsc4XB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's inspect just the \"love\" column\n",
        "idx_of_love = count_vectorizer.vocabulary_[\"love\"]\n",
        "X_train_vectorized.toarray()[:, idx_of_love]"
      ],
      "metadata": {
        "id": "rTdNdhA4n6iH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Statistical Learning\n",
        "\n",
        "Let's see of this count vector is enough to do sentiment analysis"
      ],
      "metadata": {
        "id": "e6vtukZCoYFm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# We can define our own test function\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
        "\n",
        "def test(clf):\n",
        "  clf.fit(X_train_vectorized.toarray(), y_train)\n",
        "  y_pred = clf.predict(X_test_vectorized.toarray())\n",
        "\n",
        "  print(f\"accuracy: {accuracy_score(y_true=y_test, y_pred=y_pred)}\")\n",
        "\n",
        "  return ConfusionMatrixDisplay.from_estimator(\n",
        "      clf, X_test_vectorized.toarray(), y_test,  xticks_rotation=\"vertical\"\n",
        "  )"
      ],
      "metadata": {
        "id": "OGqre19gAMQQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Naive Bayes"
      ],
      "metadata": {
        "id": "Y-6nd05OSMYO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.naive_bayes import GaussianNB\n",
        "clf = GaussianNB()\n",
        "test(clf)"
      ],
      "metadata": {
        "id": "KGsY6HLeBy4t"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Random Forest Classifier"
      ],
      "metadata": {
        "id": "CkMBsF3MSP4p"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "clf = RandomForestClassifier(max_depth=2, random_state=42)\n",
        "test(clf)"
      ],
      "metadata": {
        "id": "iKho1u5GSOdO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### Logistic Regression"
      ],
      "metadata": {
        "id": "LK8XX0BqSSzN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Logistic Regression has already been imported\n",
        "clf = LogisticRegression()\n",
        "test(clf)"
      ],
      "metadata": {
        "id": "RkdBATsESbWe"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Neural NLP\n",
        "\n",
        "## Interludium - Firth"
      ],
      "metadata": {
        "id": "sQB_BCVQ_4m8"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "This model did okaish and we could continue trying more complex models like Gradient Boosting, but eventually we have to accept that we are currently ignoring the context of words used entierly.\n",
        "\n",
        "What if we could add a hint to our word vectors where a word is used or in what context? But how would we express such an information and how would you calculate it?\n",
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "*You shall know a word by the company it keeps* (J.R. Firth 1957)\n",
        "\n",
        "The words meaning is defined by the words in its context."
      ],
      "metadata": {
        "id": "5kdo_yW7CuiG"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Word 2 Vec\n",
        "\n",
        "In 2013 researchers at Google published an idea that changed our understanding of words and directly uses the idea proposed above: The place in vector space where a word should be positioned is defined by the words typically surrounding it. Instead of a single dimension per word as in one-hot-encoding there are a shared dimensions for all words and their position can be used to infer their meaning.\n",
        "\n",
        "Such a space would be hard to craft by hand and instead is learned from a corpus. Let's see how..."
      ],
      "metadata": {
        "id": "8I80JoxK_c_H"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "[Click here to access the paper on arxiv.org.](https://arxiv.org/abs/1301.3781)"
      ],
      "metadata": {
        "id": "8PfeVxycu_VS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torch"
      ],
      "metadata": {
        "id": "85S6ZN92EmbA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "$likelihood = L(θ) = \\prod_\\limits{t=1}^{T}\\prod\\limits_{\\substack{-m\\leq j \\leq m \\\\ j \\neq 0}}P(w_{t+j}\\mid\n",
        "w_t;θ)$"
      ],
      "metadata": {
        "id": "F8owZw0eTRN4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "corpus = [\"hello\", \"world\", \"this\", \"is\", \"a\", \"signal\", \"from\", \"outer\", \"space\"]\n",
        "T = len(corpus)\n",
        "T"
      ],
      "metadata": {
        "id": "nn8SjzRTPygF"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "m = 2\n",
        "for t in range(T):\n",
        "\n",
        "  window = []\n",
        "  for j in range(-m, m+1):\n",
        "    if j == 0:\n",
        "      continue\n",
        "    if t + j < 0 or t + j >= len(corpus):\n",
        "      continue\n",
        "    window.append(corpus[t+j])\n",
        "  print(corpus[t] + \": \" +\" \".join(window))\n"
      ],
      "metadata": {
        "id": "XMhKnSeXVfTu"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### How do we calculate the $P(w_{t+j}\\mid\n",
        "w_t;θ)$?\n"
      ],
      "metadata": {
        "id": "qTnQ4zMjdSyd"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "$v_w$ vector for word $w$ if it is in the center.\n",
        "\n",
        "$u_w$ vector for word $w$ if it is in the context window.\n"
      ],
      "metadata": {
        "id": "F66QZpWoibwH"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "$softmax = P(w_{O} \\mid w_{I}) = \\frac{\\exp (u_{o}^{T} v_{i}) }{∑\\limits_{w=1}^W \\exp(u_w^Tv_{i})} $"
      ],
      "metadata": {
        "id": "BinXllx7jKne"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "unique_corpus = list(set(corpus))\n",
        "W = len(unique_corpus)\n",
        "W"
      ],
      "metadata": {
        "id": "CyKKCzxAbbn5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "v = [torch.rand(5) for _ in range(W)]\n",
        "u = [torch.rand(5) for _ in range(W)]"
      ],
      "metadata": {
        "id": "PqxrMftJmc2G"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def prob(center, context):\n",
        "    factor = torch.exp(torch.dot(u[context], v[center]))\n",
        "    normalization_constant = 0\n",
        "\n",
        "    for w in range(W):\n",
        "        normalization_constant += torch.dot(u[w], v[center])\n",
        "    return factor / normalization_constant"
      ],
      "metadata": {
        "id": "VlUICsiFvA3i"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "[Click here for a very detailed walk through word2vec](https://colab.research.google.com/drive/1_65iXT8aq_yKh6b0BkoRStPYEIMIueRP?usp=sharing\n",
        ")"
      ],
      "metadata": {
        "id": "XCiZjVzJqOEH"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## The intuition behind word2vec"
      ],
      "metadata": {
        "id": "Ipt-NIXyqYH_"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![cat_car_mouse.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "BOBlPi-kqhMV"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Using a pretrained Word2Vec model"
      ],
      "metadata": {
        "id": "NUMmi3MMDyxE"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# We can download and use a small Word2Vec model with NLTK\n",
        "# WARNING EXECUTING THIS CELL might take up to 8 minutes\n",
        "from nltk.data import find\n",
        "import gensim\n",
        "import gensim.downloader\n",
        "\n",
        "model = gensim.downloader.load('glove-twitter-200')"
      ],
      "metadata": {
        "id": "AK-7NsFBEjLW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's see the representation of \"love\"\n",
        "model[\"love\"]"
      ],
      "metadata": {
        "id": "x-oKS3rIGeYd"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# One vector has 200 dimensions\n",
        "model[\"love\"].shape"
      ],
      "metadata": {
        "id": "hdXyCFnRG1FP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Any dimensionality reduction is fine to reduce this R^200 to R^2\n",
        "from sklearn.decomposition import PCA\n",
        "\n",
        "pca = PCA(n_components=2)"
      ],
      "metadata": {
        "id": "YC96f_pBIcvp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's create some embeddings\n",
        "import numpy as np\n",
        "\n",
        "queen_embedded = model[\"queen\"]\n",
        "king_embedded = model[\"king\"]\n",
        "cat_embedded = model[\"cat\"]\n",
        "dog_embedded = model[\"dog\"]\n",
        "tree_embedded = model[\"tree\"]\n",
        "monkey_embedded = model[\"monkey\"]\n",
        "\n",
        "sub_space = np.vstack(\n",
        "    [queen_embedded, king_embedded, cat_embedded,\n",
        "     dog_embedded, tree_embedded, monkey_embedded])"
      ],
      "metadata": {
        "id": "xSwoxZ_jGHt9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# One 200 dimension vector per word ...\n",
        "sub_space.shape"
      ],
      "metadata": {
        "id": "QTELGJGCGKE9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# ... turns into one 2 dimensional vector per word\n",
        "redcued_sub_space = pca.fit_transform(sub_space)\n",
        "redcued_sub_space.shape"
      ],
      "metadata": {
        "id": "KClWLcH9GLwk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# A scatter plot is the right choice to visualise the embedding space\n",
        "import seaborn as sns\n",
        "\n",
        "\n",
        "sns.set(style='white', context='notebook', rc={'figure.figsize':(6, 6), 'figure.dpi':111})\n",
        "\n",
        "sns.scatterplot(x=redcued_sub_space[:, 0], y=redcued_sub_space[:, 1], s=150,\n",
        "                style=[\"queen\", \"king\", \"cat\", \"dog\", \"tree\", \"monkey\"],\n",
        "                hue=[\"queen\", \"king\", \"cat\", \"dog\", \"tree\", \"monkey\"])"
      ],
      "metadata": {
        "id": "bENLDSzHF-h9"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# We can reuse the preprocessing_function from above and additionally run each word through word2vec\n",
        "import numpy as np\n",
        "\n",
        "\n",
        "def embed_in_word_2_vec(text):\n",
        "    prepocessed_words = preprocessing_function(text)\n",
        "\n",
        "    representations = []\n",
        "    for word in prepocessed_words:\n",
        "        if word in model:\n",
        "            representations.append(model[word])\n",
        "\n",
        "    if representations:\n",
        "        # Why do we mean here?\n",
        "        return np.vstack(representations).mean(axis=0)\n",
        "    else:\n",
        "        return np.zeros(200)\n",
        "\n",
        "\n",
        "embed_in_word_2_vec(\"I love pizza!\").shape"
      ],
      "metadata": {
        "id": "3jiswywME9G8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_train_embedded = list(map(embed_in_word_2_vec, X_train.tolist()))\n",
        "X_test_embedded = list(map(embed_in_word_2_vec, X_test.tolist()))"
      ],
      "metadata": {
        "id": "UABgqu7VIMzb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_train_embedded_np = np.concatenate([np.expand_dims(v, axis=0) for v in X_train_embedded], axis=0)\n",
        "X_test_embedded_np = np.concatenate([np.expand_dims(v, axis=0) for v in X_test_embedded], axis=0)"
      ],
      "metadata": {
        "id": "fm6-daxCJwo2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's try the most simple neural network we have - logitstic regression\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "\n",
        "clf = LogisticRegression()\n",
        "clf.fit(X=X_train_embedded_np, y=y_train)"
      ],
      "metadata": {
        "id": "OEnzdwQdI5g6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "clf.score(X=X_test_embedded_np, y=y_test) # 🥳"
      ],
      "metadata": {
        "id": "ggxW3SWpJB4g"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Things to ponder about..."
      ],
      "metadata": {
        "id": "u_pZz6_VtCYt"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![novel.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "ZIG2bOUnqiOo"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "f9E_iLghtNh9"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}