Tensorflow chat-bot response.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "hdf5 is not supported on this machine (please install/reinstall h5py for optimal experience)\n"
     ]
    }
   ],
   "source": [
    "# things we need for NLP\n",
    "import nltk\n",
    "from nltk.stem.lancaster import LancasterStemmer\n",
    "stemmer = LancasterStemmer()\n",
    "\n",
    "# things we need for Tensorflow\n",
    "import numpy as np\n",
    "import tflearn\n",
    "import tensorflow as tf\n",
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# restore all of our data structures\n",
    "import pickle\n",
    "data = pickle.load( open( \"training_data\", \"rb\" ) )\n",
    "words = data['words']\n",
    "classes = data['classes']\n",
    "train_x = data['train_x']\n",
    "train_y = data['train_y']\n",
    "\n",
    "# import our chat-bot intents file\n",
    "import json\n",
    "with open('intents.json') as json_data:\n",
    "    intents = json.load(json_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Build neural network\n",
    "net = tflearn.input_data(shape=[None, len(train_x[0])])\n",
    "net = tflearn.fully_connected(net, 8)\n",
    "net = tflearn.fully_connected(net, 8)\n",
    "net = tflearn.fully_connected(net, len(train_y[0]), activation='softmax')\n",
    "net = tflearn.regression(net)\n",
    "\n",
    "# Define model and setup tensorboard\n",
    "model = tflearn.DNN(net, tensorboard_dir='tflearn_logs')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def clean_up_sentence(sentence):\n",
    "    # tokenize the pattern\n",
    "    sentence_words = nltk.word_tokenize(sentence)\n",
    "    # stem each word\n",
    "    sentence_words = [stemmer.stem(word.lower()) for word in sentence_words]\n",
    "    return sentence_words\n",
    "\n",
    "# return bag of words array: 0 or 1 for each word in the bag that exists in the sentence\n",
    "def bow(sentence, words, show_details=False):\n",
    "    # tokenize the pattern\n",
    "    sentence_words = clean_up_sentence(sentence)\n",
    "    # bag of words\n",
    "    bag = [0]*len(words)  \n",
    "    for s in sentence_words:\n",
    "        for i,w in enumerate(words):\n",
    "            if w == s: \n",
    "                bag[i] = 1\n",
    "                if show_details:\n",
    "                    print (\"found in bag: %s\" % w)\n",
    "\n",
    "    return(np.array(bag))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0\n",
      " 0 0 0 1 0 0 0 0 0 1 0]\n",
      "['goodbye', 'greeting', 'hours', 'mopeds', 'opentoday', 'payments', 'rental', 'thanks', 'today']\n"
     ]
    }
   ],
   "source": [
    "p = bow(\"is your shop open today?\", words)\n",
    "print (p)\n",
    "print (classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# load our saved model\n",
    "model.load('./model.tflearn')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# create a data structure to hold user context\n",
    "context = {}\n",
    "\n",
    "ERROR_THRESHOLD = 0.25\n",
    "def classify(sentence):\n",
    "    # generate probabilities from the model\n",
    "    results = model.predict([bow(sentence, words)])[0]\n",
    "    # filter out predictions below a threshold\n",
    "    results = [[i,r] for i,r in enumerate(results) if r>ERROR_THRESHOLD]\n",
    "    # sort by strength of probability\n",
    "    results.sort(key=lambda x: x[1], reverse=True)\n",
    "    return_list = []\n",
    "    for r in results:\n",
    "        return_list.append((classes[r[0]], r[1]))\n",
    "    # return tuple of intent and probability\n",
    "    return return_list\n",
    "\n",
    "def response(sentence, userID='123', show_details=False):\n",
    "    results = classify(sentence)\n",
    "    # if we have a classification then find the matching intent tag\n",
    "    if results:\n",
    "        # loop as long as there are matches to process\n",
    "        while results:\n",
    "            for i in intents['intents']:\n",
    "                # find a tag matching the first result\n",
    "                if i['tag'] == results[0][0]:\n",
    "                    # set context for this intent if necessary\n",
    "                    if 'context_set' in i:\n",
    "                        if show_details: print ('context:', i['context_set'])\n",
    "                        context[userID] = i['context_set']\n",
    "\n",
    "                    # check if this intent is contextual and applies to this user's conversation\n",
    "                    if not 'context_filter' in i or \\\n",
    "                        (userID in context and 'context_filter' in i and i['context_filter'] == context[userID]):\n",
    "                        if show_details: print ('tag:', i['tag'])\n",
    "                        # a random response from the intent\n",
    "                        return print(random.choice(i['responses']))\n",
    "\n",
    "            results.pop(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('opentoday', 0.7987914085388184)]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classify('is your shop open today?')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Our hours are 9am-9pm every day\n"
     ]
    }
   ],
   "source": [
    "response('is your shop open today?')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We accept most major credit cards\n"
     ]
    }
   ],
   "source": [
    "response('do you take cash?')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We rent Yamaha, Piaggio and Vespa mopeds\n"
     ]
    }
   ],
   "source": [
    "response('what kind of mopeds do you rent?')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Have a nice day\n"
     ]
    }
   ],
   "source": [
    "response('Goodbye, see you later')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{}"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "context"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Are you looking to rent today or later this week?\n"
     ]
    }
   ],
   "source": [
    "response('we want to rent a moped')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'123': 'rentalday'}"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show context\n",
    "context"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We accept most major credit cards\n"
     ]
    }
   ],
   "source": [
    "response('today')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('payments', 0.32852643728256226), ('today', 0.3266606032848358)]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classify('today')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "context: \n",
      "tag: greeting\n",
      "Hello, thanks for visiting\n"
     ]
    }
   ],
   "source": [
    "# clear context\n",
    "response(\"Hi there!\", show_details=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "We accept VISA, Mastercard and AMEX\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('payments', 0.32852643728256226), ('today', 0.3266606032848358)]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "response('today')\n",
    "classify('today')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Happy to help!\n"
     ]
    }
   ],
   "source": [
    "response(\"thanks, your great\")"
   ]
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}