Projects/Project_W_WikipediaLab/WikipediaLab.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Wikipedia Crawling Experiment\n",
    "\n",
    "### Idea.\n",
    "In on the Udacity's lab the instructor shared an idea that if you start at random article and keep clicking on the first link you see (excluding citations), then you will eventually get to Philosophy page. I have enhanced the algorithm to take the first link excluding references to the word origin. Otherwise, we always get to language page (typically - Latin).  \n",
    "The question I tried to explore is what are the most frequent pages we end our search at - will it be Philosophy or something else? \n",
    "\n",
    "### Tools\n",
    "For the purpose of this lab I am going to use the following libraries: \n",
    "- requests;  \n",
    "- BeutifulSoup;  \n",
    "- urllib\n",
    "- time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import requests\n",
    "import time\n",
    "import urllib\n",
    "import bs4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Search conditions\n",
    "<p>First, we need to describe the problem setup.</p>\n",
    "\n",
    "#### First link\n",
    "<p><b>The first link</b> is the first link we see in the first actual paragraph of the Wikipedia article, EXCLUDING links to references and links about the language.</p>\n",
    "\n",
    "<p>Sample fragment from Wikipedia:</p>\n",
    "<hr>\n",
    "<p>The <b>Goidelic</b> <span class=\"nowrap\"><span class=\"IPA nopopups noexcerpt\"><a href=\"/wiki/Help:IPA/English\" title=\"Help:IPA/English\">/<span style=\"border-bottom:1px dotted\"><span title=\"/ɡ/: &#39;g&#39; in &#39;guy&#39;\">ɡ</span><span title=\"/ɔɪ/: &#39;oi&#39; in &#39;choice&#39;\">ɔɪ</span><span title=\"/ˈ/: primary stress follows\">ˈ</span><span title=\"&#39;d&#39; in &#39;dye&#39;\">d</span><span title=\"/ɛ/: &#39;e&#39; in &#39;dress&#39;\">ɛ</span><span title=\"&#39;l&#39; in &#39;lie&#39;\">l</span><span title=\"/ɪ/: &#39;i&#39; in &#39;kit&#39;\">ɪ</span><span title=\"&#39;k&#39; in &#39;kind&#39;\">k</span></span>/</a></span></span> or <b>Gaelic languages</b> (<a href=\"/wiki/Irish_language\" title=\"Irish language\">Irish</a>: <i lang=\"ga\">teangacha Gaelacha</i>; <a href=\"/wiki/Scottish_Gaelic_language\" class=\"mw-redirect\" title=\"Scottish Gaelic language\">Scottish Gaelic</a>: <i lang=\"gd\">cànanan Goidhealach</i>; <a href=\"/wiki/Manx_language\" title=\"Manx language\">Manx</a>: <i lang=\"gv\">çhengaghyn Gaelgagh</i>) form one of the two groups of <a href=\"/wiki/Insular_Celtic_languages\" title=\"Insular Celtic languages\">Insular Celtic languages</a>, the other being the <a href=\"/wiki/Brittonic_languages\" title=\"Brittonic languages\">Brittonic languages</a>.<sup id=\"cite_ref-2\" class=\"reference\"><a href=\"#cite_note-2\">&#91;2&#93;</a></sup>\n",
    "</p>\n",
    "<hr>\n",
    "\n",
    "<p>In this case we should ignore the link to Irish language and \"click\" on \"Insular Celtic languages\"</p>\n",
    "\n",
    "#### Search stop conditions\n",
    "We should stop in the following cases: \n",
    "- if the chain exceeds the given length\n",
    "- if we get into a loop (e.g. Reality -> Existsence -> Reality)\n",
    "- if we get to an article with no links at all (in the main content)\n",
    "- if we reach an article on Philosophy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# start url - if we click on that link we'll get to a random Wikipedia article\n",
    "start_url = \"https://en.wikipedia.org/wiki/Special:Random\"\n",
    "target_url = \"https://en.wikipedia.org/wiki/Philosophy\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "My main modification was for the function of extracting the first link. The algorithm is as follows:  \n",
    "1) get all the html code inside the parentheses (since all language remarks are inside the parentheses);  \n",
    "2) extract all the links from the code inside of the parentheses;  \n",
    "3) when extracting the first link - loop until we meet a link that is NOT among the links obtained in step 2.\n",
    "\n",
    "The first version simply took everything after the parentheses, but this approach failed on several examples with parentheses. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_links_in_parentheses(html_element):\n",
    "    \"\"\"\n",
    "    Find all the links that are in the first parentheses in the given html element. \n",
    "    Return links in short form (as theyare in Wikipedia). \n",
    "    \"\"\"\n",
    "    elstr = html_element.prettify()\n",
    "    parenth_start = elstr.find(\"(\")\n",
    "    parenth_end = elstr.find(\")\")\n",
    "    cleaned_html = elstr[parenth_start:parenth_end+1]\n",
    "    element = bs4.BeautifulSoup(cleaned_html, \"html.parser\")\n",
    "    links = []\n",
    "    for alink in element.find_all(\"a\", recursive=False):\n",
    "        link = alink.get('href')\n",
    "        links.append(link)\n",
    "    return links"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def find_first_link(url):\n",
    "    \"\"\"Get the first link on the Wikipedia page that is not a reference\n",
    "    Core function: taken from Udacity tutorial. I've modified it to \n",
    "    search for the correct first link. \n",
    "    \"\"\"\n",
    "    response = requests.get(url)\n",
    "    html = response.text\n",
    "    soup = bs4.BeautifulSoup(html, \"html.parser\")\n",
    "\n",
    "    # This div contains the article's body\n",
    "    content_div = soup.find(id=\"mw-content-text\").find(class_=\"mw-parser-output\")\n",
    "    \n",
    "    # stores the first link found in the article, if the article contains no\n",
    "    # links this value will remain None\n",
    "    first_link = None\n",
    "    word = None\n",
    "\n",
    "    # Find all the direct children of content_div that are paragraphs\n",
    "    if content_div:\n",
    "        paragraphs = content_div.find_all(\"p\", recursive=False)\n",
    "    else:\n",
    "        return None, None\n",
    "    for element in paragraphs:\n",
    "        # Find the first anchor tag that's a direct child of a paragraph.\n",
    "        # It's important to only look at direct children, because other types\n",
    "        # of link, e.g. footnotes and pronunciation, could come before the\n",
    "        # first link to an article. Those other link types aren't direct\n",
    "        # children though, they're in divs of various classes.\n",
    "        invalid_links = get_links_in_parentheses(element)\n",
    "        # find all candidates links\n",
    "        candidates = element.find_all(\"a\", recursive=False)\n",
    "        for candidate in candidates:\n",
    "            # check if the candidate is valid (i.e. it's not a language reference)\n",
    "            if not candidate.get('href') in invalid_links:\n",
    "                article_link = candidate.get('href')\n",
    "                first_link = urllib.parse.urljoin('https://en.wikipedia.org/', article_link)\n",
    "                word = article_link.split('/')[-1]\n",
    "                break\n",
    "        if first_link:\n",
    "            break\n",
    "\n",
    "    return first_link, word"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Try this function with our example with Goidelic languages: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('https://en.wikipedia.org/wiki/Insular_Celtic_languages',\n",
       " 'Insular_Celtic_languages')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "find_first_link(\"https://en.wikipedia.org/wiki/Goidelic_languages\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def continue_crawl(search_history, target_url, max_steps):\n",
    "    \"\"\"Checks the conditions for terminating the search. Returns True if:\n",
    "    - we haven't reached Philosophy article yet\n",
    "    - the max_steps lengths was not exceeded\n",
    "    - we're not in a loop (we see the last article in the search history\n",
    "    for the first time)\n",
    "    \"\"\"\n",
    "    target = search_history[-1] == target_url\n",
    "    endless = len(search_history) > max_steps\n",
    "    loop = search_history[-1] in search_history[:-1]\n",
    "    return not target and not endless and not loop\n",
    "\n",
    "\n",
    "def find_url_chain(start_url, target_url, max_steps=25):\n",
    "    \"\"\"Returns all the links (but no more than max_steps) that are\n",
    "    accessed by clicking on the first link.\n",
    "    Core function - by Udacity. I've modified it to \n",
    "    collect the chain of words and the chain of links. \n",
    "    \"\"\"\n",
    "    start_word = start_url.split('/')[-1]\n",
    "    article_chain = [start_url]\n",
    "    word_chain = [start_word]\n",
    "    while continue_crawl(article_chain, target_url, max_steps):        \n",
    "        first_link, word = find_first_link(article_chain[-1])\n",
    "        if not first_link:            \n",
    "            print(\"We've arrived at an article with no links, aborting search!\")\n",
    "            break\n",
    "        article_chain.append(first_link)\n",
    "        word_chain.append(word)\n",
    " \n",
    "        time.sleep(1) # Slow things down so as to not hammer Wikipedia's servers\n",
    "    if len(article_chain) != len(set(article_chain)):\n",
    "        return article_chain[1:-1], word_chain[1:-1]\n",
    "    return article_chain[1:], word_chain[1:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sample links and word chains: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(['https://en.wikipedia.org/wiki/Settlor',\n",
       "  'https://en.wikipedia.org/wiki/Settlement_(trust)',\n",
       "  'https://en.wikipedia.org/wiki/Trusts_law',\n",
       "  'https://en.wikipedia.org/wiki/Fiduciary',\n",
       "  'https://en.wikipedia.org/wiki/Law',\n",
       "  'https://en.wikipedia.org/wiki/System',\n",
       "  'https://en.wikipedia.org/wiki/Latin',\n",
       "  'https://en.wikipedia.org/wiki/Classical_language',\n",
       "  'https://en.wikipedia.org/wiki/Language',\n",
       "  'https://en.wikipedia.org/wiki/Communication',\n",
       "  'https://en.wikipedia.org/wiki/Meaning_(semiotics)',\n",
       "  'https://en.wikipedia.org/wiki/Semiotics',\n",
       "  'https://en.wikipedia.org/wiki/Meaning-making',\n",
       "  'https://en.wikipedia.org/wiki/Psychology',\n",
       "  'https://en.wikipedia.org/wiki/Science',\n",
       "  'https://en.wikipedia.org/wiki/Knowledge',\n",
       "  'https://en.wikipedia.org/wiki/Fact',\n",
       "  'https://en.wikipedia.org/wiki/Reality',\n",
       "  'https://en.wikipedia.org/wiki/Existence'],\n",
       " ['Settlor',\n",
       "  'Settlement_(trust)',\n",
       "  'Trusts_law',\n",
       "  'Fiduciary',\n",
       "  'Law',\n",
       "  'System',\n",
       "  'Latin',\n",
       "  'Classical_language',\n",
       "  'Language',\n",
       "  'Communication',\n",
       "  'Meaning_(semiotics)',\n",
       "  'Semiotics',\n",
       "  'Meaning-making',\n",
       "  'Psychology',\n",
       "  'Science',\n",
       "  'Knowledge',\n",
       "  'Fact',\n",
       "  'Reality',\n",
       "  'Existence'])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "find_url_chain('https://en.wikipedia.org/wiki/Special:Random', \"https://en.wikipedia.org/wiki/Philosophy\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Experiment\n",
    "Now we can carry out an experiment. We will run the whole process for 20 different articles. I've added try-catch block in case some of these searches fails. The reason is that we might encounter some unpredictable situation (since it's impossible to predict every possible format of Wikipedia article), and in this case we would like to simply drop this article. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "def get_url_chains(n_articles=3, target_url=\"https://en.wikipedia.org/wiki/Philosophy\"):\n",
    "    all_chains = []\n",
    "    all_word_chains = []    \n",
    "    for i in range(n_articles):\n",
    "        try:\n",
    "            chain, word_chain = find_url_chain(\"https://en.wikipedia.org/wiki/Special:Random\", target_url)\n",
    "            all_chains.append(chain)\n",
    "            all_word_chains.append(word_chain)\n",
    "        except e:\n",
    "            e = sys.exc_info()[0]\n",
    "            print(e)            \n",
    "    return all_chains, all_word_chains        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We've arrived at an article with no links, aborting search!\n",
      "We've arrived at an article with no links, aborting search!\n",
      "We've arrived at an article with no links, aborting search!\n"
     ]
    }
   ],
   "source": [
    "all_chains, all_word_chains = get_url_chains(100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Vizualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_links = [y for x in all_chains for y in x]\n",
    "all_words = [y for x in all_word_chains for y in x]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Knowledge', 83),\n",
       " ('Fact', 83),\n",
       " ('Reality', 82),\n",
       " ('Existence', 82),\n",
       " ('Science', 74),\n",
       " ('Psychology', 39),\n",
       " ('Natural_science', 23),\n",
       " ('Branches_of_science', 23),\n",
       " ('Communication', 22),\n",
       " ('Meaning_(semiotics)', 22)]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from collections import Counter\n",
    "word_counts = Counter(all_words).most_common()[:20]\n",
    "word_counts[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#xs = list(word_counts.keys())\n",
    "x, y = zip(*word_counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(14, 8))\n",
    "ax = fig.add_subplot(111)\n",
    "rects1 = ax.bar(x, y, align = 'center')\n",
    "ax.set_xticks(x)\n",
    "ax.set_xticklabels(x, rotation=90)\n",
    "ax.set_title(\"Number of occurences of Wikipedia articles in 100 random first-link-click runs\", size=16)\n",
    "ax.set_xlabel(\"Article Names\", size=14)\n",
    "ax.set_ylabel(\"Number of occurences in all chains\",size=14)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}