How to build a bot.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Infobot built using NLTK-Chatbot\n",
    "A simple bot that is able to search Wikipedia for information.\n",
    "NLTK (Natural Language Toolkit)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Things used to Build"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "*1. NLTK - Chatbot* "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Chatbot is built with RegEx (Regular expression) and NLU (Natural Language Understanding)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "*2. Django* "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Django is a free and open-source web framework, written in Python, which follows the model-view-template (MVT) architectural pattern"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "*3. Facebook Messengerbot* "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Python client for Facebook Messenger Platform API"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Initial imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "from chatbot import Chat, register_call\n",
    "import wikipedia"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Wikipedia Summary extraction\n",
    "Integrates the wikipedia API to the chatbot program by registering function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "@register_call(\"whoIs\")\n",
    "def whoIs(query,session_id=\"general\"):\n",
    "    try:\n",
    "        return wikipedia.summary(query)\n",
    "    except Exception:\n",
    "        for newquery in wikipedia.search(query):\n",
    "            try:\n",
    "                return wikipedia.summary(newquery)\n",
    "            except Exception:\n",
    "                pass\n",
    "    return \"I don't know about \"+query"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Create template File"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Building the knowledge base for the chat\n",
    "The responses and actions to be taken for a user's message is stored in a knowledge base"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Handling a message after asking for his/her name"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "```\n",
    "{% block %}\n",
    "    {% client %}((I('| a)m|my name is) )?(.*){% endclient %}\n",
    "    {% prev %}(.*)Can you please tell me your name{% endprev %}\n",
    "    {% response %}Thank you {name:%4}{% endresponse %}\n",
    "{% endblock %}\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### User asking the bot for his/her name"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "```\n",
    "{% block %}\n",
    "    {% client %}Do you remember my name{% endclient %}\n",
    "    {% response %}{% if {name} %}Yes I do {name}{% else %}No,{% chat what is my name %}{% endif %}{% endresponse %}\n",
    "{% endblock %}\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Handling message that request for some information"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "```\n",
    "{% block %}\n",
    "    {% client %}(Do you know about|what is|who is|tell me about)(.*){% endclient %}\n",
    "    {% response %}{% call whoIs:%2 %}{% endresponse %}\n",
    "{% endblock %}\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Multiple variations for responses for a single type of message"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "```\n",
    "{% block %}\n",
    "    {% client %}Hello(.*){% endclient %}\n",
    "    {% response %}Hello... I'm glad you could drop by today.{% endresponse %}\n",
    "    {% response %}Hi there... how are you today?{% endresponse %}\n",
    "    {% response %}Hello, how are you feeling today?{% endresponse %}\n",
    "{% endblock %}\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Specifying an initial starter message\n",
    "For this example the bot starts the conversation. You can also have the user start the conversation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "firstQuestion=\"Hi, how are you?\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## The conversation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hi, how are you?\n"
     ]
    }
   ],
   "source": [
    "Chat(\"demo.template\").converse(firstQuestion)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}