diff --git a/Twitter_Sentiment_Analysis.ipynb b/Twitter_Sentiment_Analysis.ipynb
index 7a3345e..716a085 100644
--- a/Twitter_Sentiment_Analysis.ipynb
+++ b/Twitter_Sentiment_Analysis.ipynb
@@ -107,10 +107,10 @@
"metadata": {
"id": "pXSrmmZr7orS",
"colab_type": "code",
- "outputId": "0528d36c-3b50-4ce8-dce4-a5eda7b27679",
+ "outputId": "4d3612ec-f2ba-4e61-9703-eef93a53265a",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 377
+ "height": 394
}
},
"source": [
@@ -163,33 +163,40 @@
"from sklearn.model_selection import GridSearchCV\n",
"from sklearn.externals import joblib"
],
- "execution_count": 3,
+ "execution_count": 1,
"outputs": [
{
"output_type": "stream",
"text": [
"Requirement already satisfied: tweepy in /usr/local/lib/python3.6/dist-packages (3.6.0)\n",
- "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tweepy) (1.12.0)\n",
- "Requirement already satisfied: PySocks>=1.5.7 in /usr/local/lib/python3.6/dist-packages (from tweepy) (1.7.1)\n",
"Requirement already satisfied: requests>=2.11.1 in /usr/local/lib/python3.6/dist-packages (from tweepy) (2.23.0)\n",
"Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.6/dist-packages (from tweepy) (1.3.0)\n",
- "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests>=2.11.1->tweepy) (2.9)\n",
+ "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tweepy) (1.12.0)\n",
+ "Requirement already satisfied: PySocks>=1.5.7 in /usr/local/lib/python3.6/dist-packages (from tweepy) (1.7.1)\n",
"Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests>=2.11.1->tweepy) (2020.4.5.1)\n",
"Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests>=2.11.1->tweepy) (3.0.4)\n",
"Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests>=2.11.1->tweepy) (1.24.3)\n",
+ "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests>=2.11.1->tweepy) (2.9)\n",
"Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.6/dist-packages (from requests-oauthlib>=0.7.0->tweepy) (3.1.0)\n",
"[nltk_data] Downloading package punkt to /root/nltk_data...\n",
- "[nltk_data] Package punkt is already up-to-date!\n",
+ "[nltk_data] Unzipping tokenizers/punkt.zip.\n",
"[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
- "[nltk_data] Package stopwords is already up-to-date!\n",
+ "[nltk_data] Unzipping corpora/stopwords.zip.\n",
"[nltk_data] Downloading package wordnet to /root/nltk_data...\n",
- "[nltk_data] Package wordnet is already up-to-date!\n",
+ "[nltk_data] Unzipping corpora/wordnet.zip.\n",
"[nltk_data] Downloading package averaged_perceptron_tagger to\n",
"[nltk_data] /root/nltk_data...\n",
- "[nltk_data] Package averaged_perceptron_tagger is already up-to-\n",
- "[nltk_data] date!\n"
+ "[nltk_data] Unzipping taggers/averaged_perceptron_tagger.zip.\n"
],
"name": "stdout"
+ },
+ {
+ "output_type": "stream",
+ "text": [
+ "/usr/local/lib/python3.6/dist-packages/sklearn/externals/joblib/__init__.py:15: FutureWarning: sklearn.externals.joblib is deprecated in 0.21 and will be removed in 0.23. Please import this functionality directly from joblib, which can be installed with: pip install joblib. If this warning is raised when loading pickled models, you may need to re-serialize those models with scikit-learn 0.21+.\n",
+ " warnings.warn(msg, category=FutureWarning)\n"
+ ],
+ "name": "stderr"
}
]
},
@@ -317,17 +324,17 @@
"metadata": {
"id": "hQDprSKllSNL",
"colab_type": "code",
- "outputId": "e0447214-0427-4a1a-b526-4bacb6f0274d",
+ "outputId": "f3cb8325-5421-45c0-f9cc-60f0a23e3278",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 391
+ "height": 204
}
},
"source": [
"data1 = pd.read_csv('https://github.com/TharinduMunasinge/Twitter-Sentiment-Analysis/raw/master/DataSet/FinalizedFull.csv').rename(columns={'tweet':'text'})\n",
"data1.head()"
],
- "execution_count": 0,
+ "execution_count": 3,
"outputs": [
{
"output_type": "execute_result",
@@ -397,77 +404,7 @@
"metadata": {
"tags": []
},
- "execution_count": 23
- },
- {
- "output_type": "execute_result",
- "data": {
- "text/html": [
- "
\n",
- "\n",
- "
\n",
- " \n",
- " \n",
- " | \n",
- " text | \n",
- " senti | \n",
- "
\n",
- " \n",
- " \n",
- " \n",
- " 0 | \n",
- " @united UA5396 can wait for me. I'm on the gro... | \n",
- " 0 | \n",
- "
\n",
- " \n",
- " 1 | \n",
- " I hate Time Warner! Soooo wish I had Vios. Can... | \n",
- " 0 | \n",
- "
\n",
- " \n",
- " 2 | \n",
- " @united Oh, we are sure it's not planned, but ... | \n",
- " 0 | \n",
- "
\n",
- " \n",
- " 3 | \n",
- " Tom Shanahan's latest column on SDSU and its N... | \n",
- " 2 | \n",
- "
\n",
- " \n",
- " 4 | \n",
- " Found the self driving car!! /IWo3QSvdu2 | \n",
- " 2 | \n",
- "
\n",
- " \n",
- "
\n",
- "
"
- ],
- "text/plain": [
- " text senti\n",
- "0 @united UA5396 can wait for me. I'm on the gro... 0\n",
- "1 I hate Time Warner! Soooo wish I had Vios. Can... 0\n",
- "2 @united Oh, we are sure it's not planned, but ... 0\n",
- "3 Tom Shanahan's latest column on SDSU and its N... 2\n",
- "4 Found the self driving car!! /IWo3QSvdu2 2"
- ]
- },
- "metadata": {
- "tags": []
- },
- "execution_count": 34
+ "execution_count": 3
}
]
},
@@ -476,24 +413,21 @@
"metadata": {
"id": "sOCWOfJFPLUd",
"colab_type": "code",
- "outputId": "d5c4726b-abde-40fb-cebe-64de9bdedaf3",
+ "outputId": "a904e268-1d20-44d7-ac94-c09b648335a0",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 119
+ "height": 68
}
},
"source": [
"# Check the balance of the dataset\n",
"counts1 = print_balance(data1)"
],
- "execution_count": 0,
+ "execution_count": 4,
"outputs": [
{
"output_type": "stream",
"text": [
- "This dataset contains 377 negative tweets (37.8%)\n",
- "This dataset contains 239 neutral tweets (24.0%)\n",
- "This dataset contains 381 positive tweets (38.2%)\n",
"This dataset contains 377 negative tweets (37.8%)\n",
"This dataset contains 239 neutral tweets (24.0%)\n",
"This dataset contains 381 positive tweets (38.2%)\n"
@@ -527,10 +461,10 @@
"metadata": {
"id": "Ks9cEN71BvyK",
"colab_type": "code",
- "outputId": "29eaf93c-b1a0-42fe-e0c5-d2118ff47f5e",
+ "outputId": "fc799ee0-2da7-4b3d-82fb-38f6066fbee5",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 391
+ "height": 204
}
},
"source": [
@@ -545,7 +479,7 @@
"\n",
"data2.head()"
],
- "execution_count": 0,
+ "execution_count": 5,
"outputs": [
{
"output_type": "execute_result",
@@ -615,77 +549,7 @@
"metadata": {
"tags": []
},
- "execution_count": 25
- },
- {
- "output_type": "execute_result",
- "data": {
- "text/html": [
- "\n",
- "\n",
- "
\n",
- " \n",
- " \n",
- " | \n",
- " senti | \n",
- " text | \n",
- "
\n",
- " \n",
- " \n",
- " \n",
- " 0 | \n",
- " 4 | \n",
- " Now all @Apple has to do is get swype on the i... | \n",
- "
\n",
- " \n",
- " 1 | \n",
- " 4 | \n",
- " @Apple will be adding more carrier support to ... | \n",
- "
\n",
- " \n",
- " 2 | \n",
- " 4 | \n",
- " Hilarious @youtube video - guy does a duet wit... | \n",
- "
\n",
- " \n",
- " 3 | \n",
- " 4 | \n",
- " @RIM you made it too easy for me to switch to ... | \n",
- "
\n",
- " \n",
- " 4 | \n",
- " 4 | \n",
- " I just realized that the reason I got into twi... | \n",
- "
\n",
- " \n",
- "
\n",
- "
"
- ],
- "text/plain": [
- " senti text\n",
- "0 4 Now all @Apple has to do is get swype on the i...\n",
- "1 4 @Apple will be adding more carrier support to ...\n",
- "2 4 Hilarious @youtube video - guy does a duet wit...\n",
- "3 4 @RIM you made it too easy for me to switch to ...\n",
- "4 4 I just realized that the reason I got into twi..."
- ]
- },
- "metadata": {
- "tags": []
- },
- "execution_count": 36
+ "execution_count": 5
}
]
},
@@ -694,24 +558,21 @@
"metadata": {
"id": "xHT5w5apSblQ",
"colab_type": "code",
- "outputId": "c0c924e6-b291-41ae-c4f1-e751972ae8d4",
+ "outputId": "50d3f8c1-6699-4ebd-e27a-8ac3e8c218f7",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 119
+ "height": 68
}
},
"source": [
"# Check the balance of the dataset\n",
"counts2 = print_balance(data2)"
],
- "execution_count": 0,
+ "execution_count": 6,
"outputs": [
{
"output_type": "stream",
"text": [
- "This dataset contains 572 negative tweets (16.7%)\n",
- "This dataset contains 2333 neutral tweets (68.1%)\n",
- "This dataset contains 519 positive tweets (15.2%)\n",
"This dataset contains 572 negative tweets (16.7%)\n",
"This dataset contains 2333 neutral tweets (68.1%)\n",
"This dataset contains 519 positive tweets (15.2%)\n"
@@ -737,10 +598,10 @@
"metadata": {
"id": "nILYW_U3UgY0",
"colab_type": "code",
- "outputId": "bbe4274d-bbe4-45ed-b98b-73ae65ad23cf",
+ "outputId": "028ef27e-03ef-45d5-95d9-d36ee3ce4977",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 119
+ "height": 68
}
},
"source": [
@@ -751,14 +612,11 @@
"# Check the balance of the dataset\n",
"train_counts = print_balance(train)"
],
- "execution_count": 0,
+ "execution_count": 7,
"outputs": [
{
"output_type": "stream",
"text": [
- "This dataset contains 949 negative tweets (21.5%)\n",
- "This dataset contains 2572 neutral tweets (58.2%)\n",
- "This dataset contains 900 positive tweets (20.4%)\n",
"This dataset contains 949 negative tweets (21.5%)\n",
"This dataset contains 2572 neutral tweets (58.2%)\n",
"This dataset contains 900 positive tweets (20.4%)\n"
@@ -796,10 +654,10 @@
"metadata": {
"id": "-xAwNoKD8okU",
"colab_type": "code",
- "outputId": "842f5601-13f7-4bcd-9d63-47beb54e7671",
+ "outputId": "8bf2683d-91af-4ec8-d2a8-56ce5bd4d0da",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 119
+ "height": 68
}
},
"source": [
@@ -812,14 +670,11 @@
"# Create train and test sets\n",
"X_train, X_test, y_train, y_test = train_test_split(train_adj.text.values, train_adj.senti.values, random_state=637)"
],
- "execution_count": 0,
+ "execution_count": 8,
"outputs": [
{
"output_type": "stream",
"text": [
- "This dataset contains 900 negative tweets (28.6%)\n",
- "This dataset contains 1350 neutral tweets (42.9%)\n",
- "This dataset contains 900 positive tweets (28.6%)\n",
"This dataset contains 900 negative tweets (28.6%)\n",
"This dataset contains 1350 neutral tweets (42.9%)\n",
"This dataset contains 900 positive tweets (28.6%)\n"
@@ -857,10 +712,10 @@
"metadata": {
"id": "zwvToCqPIX7M",
"colab_type": "code",
- "outputId": "a2ceee8c-df8d-4216-90e1-fc8400b544b2",
+ "outputId": "1ba76319-d6c7-4abb-c6a5-2931e8abd500",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 323
+ "height": 170
}
},
"source": [
@@ -880,20 +735,11 @@
"X_test_bal = test_bal.text.values\n",
"y_test_bal = test_bal.senti.values"
],
- "execution_count": 0,
+ "execution_count": 10,
"outputs": [
{
"output_type": "stream",
"text": [
- "\u001b[1mComposition of unbalanced test set:\u001b[0m\n",
- "This dataset contains 111 negative tweets (20.2%)\n",
- "This dataset contains 327 neutral tweets (59.1%)\n",
- "This dataset contains 114 positive tweets (20.7%)\n",
- "\n",
- "\u001b[1mComposition of balanced test set:\u001b[0m\n",
- "This dataset contains 114 negative tweets (33.3%)\n",
- "This dataset contains 114 neutral tweets (33.3%)\n",
- "This dataset contains 114 positive tweets (33.3%)\n",
"\u001b[1mComposition of unbalanced test set:\u001b[0m\n",
"This dataset contains 111 negative tweets (20.2%)\n",
"This dataset contains 327 neutral tweets (59.1%)\n",
@@ -1247,7 +1093,7 @@
" print('Calculating uncertainty for aggregate sentiment scores...')\n",
" self.agg_errs, self.agg_true_means, self.agg_pred_means = self.agg_sent_score(X, y,\n",
" n_bootstraps=1000,\n",
- " bootstrap_size=0.5,\n",
+ " bootstrap_size=1000,\n",
" verbose=True)\n",
" self.agg_me = self.agg_errs.mean() # Mean average error in aggregated sentiment scores\n",
" self.agg_unc = model.agg_errs.std()*2 # Uncertainty in aggregated sentiment scores (95% confidence)\n",
@@ -1275,7 +1121,8 @@
"\n",
" def predict_agg(self, X, verbose=False):\n",
" agg_sent = ((self.grid.predict(X) - 2) / 2).mean()\n",
- " if verbose: print('\\n(scale from -1 to 1) Aggregated sentiment score: {0:.2f} \\u00B1 {1:.2f}\\n'.format(agg_sent, self.agg_unc))\n",
+ " agg_unc = self.agg_unc * np.sqrt(1000 / len(X))\n",
+ " if verbose: print('(scale from -1 to 1) Aggregated sentiment score: {0:.2f} \\u00B1 {1:.2f}\\n'.format(agg_sent, agg_unc))\n",
" return agg_sent, self.agg_unc\n",
"\n",
"\n",
@@ -1377,14 +1224,13 @@
" return fig\n",
"\n",
"\n",
- " def agg_sent_score(self, X, y, n_bootstraps=1000, bootstrap_size=0.5, verbose=False):\n",
+ " def agg_sent_score(self, X, y, n_bootstraps=1000, bootstrap_size=1000, verbose=False):\n",
" np.random.seed(637)\n",
- " size = int(len(X) * bootstrap_size)\n",
" true_means, pred_means, errs = np.ones(n_bootstraps), np.ones(n_bootstraps), np.ones(n_bootstraps)\n",
" for i in range(n_bootstraps):\n",
" if verbose:\n",
" if not (i+1)%(n_bootstraps//5): print('{0:.0f}% done'.format((i+1)/(n_bootstraps//100)))\n",
- " ind = np.random.randint(0, len(X), size)\n",
+ " ind = np.random.randint(0, len(X), int(bootstrap_size))\n",
" X_boot = X[ind]\n",
" y_boot = y[ind]\n",
" y_pred = self.grid.predict(X_boot)\n",
@@ -1401,7 +1247,7 @@
" return errs, true_means, pred_means\n",
"\n",
"\n",
- " def agg_sent_hist(self, X, y, n_bootstraps=1000, bootstrap_size=0.5, verbose=False):\n",
+ " def agg_sent_hist(self, X, y, n_bootstraps=1000, bootstrap_size=1000, verbose=False):\n",
" errs, true_means, pred_means = self.agg_sent_score(X, y, n_bootstraps, bootstrap_size, verbose)\n",
" fig, ax = plt.subplots(figsize=(7, 7))\n",
" ax.hist(errs)\n",
@@ -1412,7 +1258,7 @@
" return fig\n",
"\n",
"\n",
- " def agg_sent_plot(self, X, y, n_bootstraps=1000, bootstrap_size=0.5, verbose=False):\n",
+ " def agg_sent_plot(self, X, y, n_bootstraps=1000, bootstrap_size=1000, verbose=False):\n",
" errs, true_means, pred_means = self.agg_sent_score(X, y, n_bootstraps, bootstrap_size, verbose)\n",
" fig, ax = plt.subplots(figsize=(7, 7))\n",
" ax.scatter(true_means, pred_means)\n",
@@ -1484,26 +1330,27 @@
"metadata": {
"id": "qsH0VP_pW-OR",
"colab_type": "code",
- "outputId": "a467e4ea-af07-4377-d4d0-3b9200a9352d",
+ "outputId": "ecfcdea1-60e9-46ec-cec8-b503a753e907",
"colab": {
"base_uri": "https://localhost:8080/",
- "height": 445
+ "height": 394
}
},
"source": [
- "model = TwitterSentimentModel()\n",
- "model.fit(X_train, y_train, classifier='all')\n",
- "model.set_uncertainty(X_test, y_test)\n",
+ "if False:\n",
+ " model = TwitterSentimentModel()\n",
+ " model.fit(X_train, y_train, classifier='SGD')\n",
+ " model.set_uncertainty(X_test, y_test)\n",
"\n",
- "with open(\"trained_model.pickle\", \"wb\") as f:\n",
- " pickle.dump(model, f)"
+ " with open(\"trained_model.pickle\", \"wb\") as f:\n",
+ " pickle.dump(model, f)"
],
"execution_count": 0,
"outputs": [
{
"output_type": "stream",
"text": [
- "Fitting 10 folds for each of 132 candidates, totalling 1320 fits\n"
+ "Fitting 10 folds for each of 24 candidates, totalling 240 fits\n"
],
"name": "stdout"
},
@@ -1511,12 +1358,9 @@
"output_type": "stream",
"text": [
"[Parallel(n_jobs=-1)]: Using backend LokyBackend with 2 concurrent workers.\n",
- "[Parallel(n_jobs=-1)]: Done 46 tasks | elapsed: 55.7s\n",
- "[Parallel(n_jobs=-1)]: Done 196 tasks | elapsed: 2.3min\n",
- "[Parallel(n_jobs=-1)]: Done 446 tasks | elapsed: 2.8min\n",
- "[Parallel(n_jobs=-1)]: Done 796 tasks | elapsed: 3.4min\n",
- "[Parallel(n_jobs=-1)]: Done 1246 tasks | elapsed: 5.7min\n",
- "[Parallel(n_jobs=-1)]: Done 1320 out of 1320 | elapsed: 6.0min finished\n"
+ "[Parallel(n_jobs=-1)]: Done 46 tasks | elapsed: 6.8s\n",
+ "[Parallel(n_jobs=-1)]: Done 196 tasks | elapsed: 28.8s\n",
+ "[Parallel(n_jobs=-1)]: Done 240 out of 240 | elapsed: 35.5s finished\n"
],
"name": "stderr"
},
@@ -1538,7 +1382,7 @@
"60% done\n",
"80% done\n",
"100% done\n",
- "Done! Aggregate sentiment uncertainty = 0.0494\n"
+ "Done! Aggregate sentiment uncertainty = 0.0353\n"
],
"name": "stdout"
}
@@ -1573,7 +1417,7 @@
"metadata": {
"id": "WTrBk5ecFw4N",
"colab_type": "code",
- "outputId": "f7e00bd4-2f0d-412b-f53d-127c8e2692dd",
+ "outputId": "ea182b06-2b89-44c3-864d-13ab768a313b",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 51
@@ -1601,7 +1445,7 @@
"metadata": {
"id": "_1skobezDS2r",
"colab_type": "code",
- "outputId": "b21d7ab2-802d-4088-f43c-c60f881e8401",
+ "outputId": "b8b94f47-e497-452d-9236-72f74f92f76f",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 466
@@ -1633,7 +1477,7 @@
"metadata": {
"id": "s5X5AWId2jWp",
"colab_type": "code",
- "outputId": "c21ad1d3-a520-48ce-fc2a-b6530941e990",
+ "outputId": "4cbf5fb3-5e2d-4d4e-a3ae-dd1d70f8263d",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 487
@@ -1649,14 +1493,14 @@
{
"output_type": "stream",
"text": [
- "Mean aggregate score error: 0.00407\n"
+ "Mean aggregate score error: 0.00379\n"
],
"name": "stdout"
},
{
"output_type": "display_data",
"data": {
- "image/png": "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\n",
+ "image/png": "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\n",
"text/plain": [
"