{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport nltk\nimport string\nfrom nltk.tokenize import word_tokenize\nfrom nltk.corpus import stopwords\nfrom nltk.stem import WordNetLemmatizer\nimport pandas as pd \nimport seaborn as sns\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-01-07T16:50:48.265927Z","iopub.execute_input":"2022-01-07T16:50:48.266412Z","iopub.status.idle":"2022-01-07T16:50:50.155297Z","shell.execute_reply.started":"2022-01-07T16:50:48.266312Z","shell.execute_reply":"2022-01-07T16:50:50.154032Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"[![image.png](attachment:0ea6dd08-089d-42cd-8a4b-e5de07edbf81.png)](http://)","metadata":{},"attachments":{"0ea6dd08-089d-42cd-8a4b-e5de07edbf81.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"# Kết quả chạy trên kaggle:","metadata":{}},{"cell_type":"markdown","source":"# Mô tả bài toán\n> ***Yêu cầu đặt ra là với các câu hỏi trên Quora có phải là câu hỏi toxic hay không?***\n* Input: Các câu dưới dạng text.\n* Output: 1(Là câu hỏi toxic) or 0(Là câu bình thường) ?\n","metadata":{}},{"cell_type":"markdown","source":"# Tiến hành bài toán theo:\n**1. Xử lý dữ liệu**\n> Vì đầu vào là text nên chúng ta cần xử lý trước khi train\n\n**2. Training dữ liệu với mô hình Logistic Regression**\n\n**3. Đánh giá và kết luận**","metadata":{}},{"cell_type":"markdown","source":"# 1. Xử lý dữ liệu","metadata":{}},{"cell_type":"markdown","source":"**Đọc dữ liệu đầu vào gồm 3 tập: **\n\n* Sample submission\n* Train\n* Test\n","metadata":{}},{"cell_type":"code","source":"df_sample_sub = pd.read_csv('/kaggle/input/quora-insincere-questions-classification/sample_submission.csv')\ndf_train = pd.read_csv('/kaggle/input/quora-insincere-questions-classification/train.csv')\ndf_test = pd.read_csv('/kaggle/input/quora-insincere-questions-classification/test.csv')","metadata":{"execution":{"iopub.status.busy":"2022-01-07T16:50:50.157152Z","iopub.execute_input":"2022-01-07T16:50:50.157534Z","iopub.status.idle":"2022-01-07T16:50:57.700478Z","shell.execute_reply.started":"2022-01-07T16:50:50.157495Z","shell.execute_reply":"2022-01-07T16:50:57.699105Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Dữ liệu sẽ hiển thị dạng**","metadata":{}},{"cell_type":"code","source":"# Lấy dữ liệu tập sample_sub\ndf_sample_sub.head()","metadata":{"execution":{"iopub.status.busy":"2022-01-07T16:51:09.778509Z","iopub.execute_input":"2022-01-07T16:51:09.778945Z","iopub.status.idle":"2022-01-07T16:51:09.791234Z","shell.execute_reply.started":"2022-01-07T16:51:09.778911Z","shell.execute_reply":"2022-01-07T16:51:09.790116Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Lấy dữ liệu tập train\ndf_train.head()","metadata":{"execution":{"iopub.status.busy":"2022-01-07T16:51:18.652042Z","iopub.execute_input":"2022-01-07T16:51:18.65249Z","iopub.status.idle":"2022-01-07T16:51:18.664207Z","shell.execute_reply.started":"2022-01-07T16:51:18.652452Z","shell.execute_reply":"2022-01-07T16:51:18.663304Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Dữ liệu vào gồm các question_text là dạng text. -> Ta sẽ phải xử lý dấu câu, các \"stopword\", từ đồng nghĩa để có thể dễ dàng hơn cho việc train.**","metadata":{}},{"cell_type":"code","source":"# Thông tin tập train\ndf_train.info()","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:42:44.09382Z","iopub.execute_input":"2022-01-07T15:42:44.094273Z","iopub.status.idle":"2022-01-07T15:42:44.382161Z","shell.execute_reply.started":"2022-01-07T15:42:44.094226Z","shell.execute_reply":"2022-01-07T15:42:44.380896Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Lượng dữ liệu \ndf_train.target.value_counts()","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:42:44.383556Z","iopub.execute_input":"2022-01-07T15:42:44.383933Z","iopub.status.idle":"2022-01-07T15:42:44.402728Z","shell.execute_reply.started":"2022-01-07T15:42:44.383898Z","shell.execute_reply":"2022-01-07T15:42:44.401592Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Đối với tập dữ liệu Train**\n\n**Có 1225312 câu hỏi bình thường và 80810 câu hỏi Toxic**\n\n**-> Tỉ lệ chênh lệch lớn và các câu hỏi toxic vẫn xuất hiện khá nhiều chiếm tới 80810 câu trong tập dữ liệu.**\n","metadata":{}},{"cell_type":"markdown","source":"**Sử dụng Natural Language Toolkit để xử lý ngôn ngữ tự nhiên, nhanh chóng và có tác dụng làm sạch dữ liệu, xử lý dữ liệu đầu vào cho các thuật toán Machine Learning và giúp ta xử lý các stopwords - là các từ có tần số xuất hiện nhiều nhưng mang lại giá trị ý nghĩa không khác mấy so với khi ta bỏ đi.\nVí dụ: my, me, ...**","metadata":{}},{"cell_type":"code","source":"nltk.download('stopwords')\nnltk_stopwords = stopwords.words('english')\nwordnet_lemmatizer = WordNetLemmatizer()\n\ndef lemSentence(sentence):\n    token_words = word_tokenize(sentence)\n    lem_sentence = []\n    for word in token_words:\n        lem_sentence.append(wordnet_lemmatizer.lemmatize(word, pos=\"v\"))\n        lem_sentence.append(\" \")\n    return \"\".join(lem_sentence)","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:42:44.829241Z","iopub.execute_input":"2022-01-07T15:42:44.829576Z","iopub.status.idle":"2022-01-07T15:43:04.895153Z","shell.execute_reply.started":"2022-01-07T15:42:44.829544Z","shell.execute_reply":"2022-01-07T15:43:04.893935Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Tiến hành clean dữ liệu**","metadata":{}},{"cell_type":"code","source":"def clean(message, lem = True):\n    # Loại bỏ dấu câu (ví dụ: các dấu !\"#$%&'()*+, -./:;<=>?@[\\]^_`{|}~)\n    message = message.translate(str.maketrans('', '', string.punctuation))\n    \n    # Loại bỏ số\n    message = message.translate(str.maketrans('', '', string.digits))\n    \n    # Loại bỏ \"stopwords\"\n    message = [word for word in word_tokenize(message) if not word.lower() in nltk_stopwords]\n    message = ' '.join(message)\n    \n    if lem:\n        message = lemSentence(message)\n    \n    return message","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:43:04.897099Z","iopub.execute_input":"2022-01-07T15:43:04.897397Z","iopub.status.idle":"2022-01-07T15:43:04.903179Z","shell.execute_reply.started":"2022-01-07T15:43:04.89737Z","shell.execute_reply":"2022-01-07T15:43:04.902206Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Clean các câu hỏi trong tập train\ndf_train['question_text_cleaned'] = df_train.question_text.apply(lambda x: clean(x, True))","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:43:04.904584Z","iopub.execute_input":"2022-01-07T15:43:04.904958Z","iopub.status.idle":"2022-01-07T15:52:14.779123Z","shell.execute_reply.started":"2022-01-07T15:43:04.904925Z","shell.execute_reply":"2022-01-07T15:52:14.778099Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Bảng 20 dữ liệu sau khi tiến hành làm sạch sẽ hiển thị như sau:**","metadata":{}},{"cell_type":"code","source":"df_train.head(10)","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:54:16.01697Z","iopub.execute_input":"2022-01-07T15:54:16.017386Z","iopub.status.idle":"2022-01-07T15:54:16.033847Z","shell.execute_reply.started":"2022-01-07T15:54:16.017347Z","shell.execute_reply":"2022-01-07T15:54:16.032799Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 2. Training dữ liệu với mô hình hồi quy logistic (Logistic Regression)\n\n**> Tổng quan về bài toán được trình bày trên link notion đã có ở trên lớp: https://tricky-tax-444.notion.site/5acbc01b2a924fe6818d046339a68d23?v=6bc56f801d7c4ce1a1eefd4b60ef57cf&p=cf099245442c461aa36305456a17183e\nvà https://machinelearningcoban.com/2017/01/27/logisticregression/**","metadata":{}},{"cell_type":"markdown","source":"**Tuy nhiên sau khi đó ta vẫn phải mã hoá văn bản. Đầu vào cần được mã hóa dưới dạng số nguyên hoặc giá trị dấu phẩy động, để sử dụng làm đầu vào trong thuật toán học máy. Quá trình này được gọi là vecto hoá.**\n\n**-> Ta sẽ sử dụng CountVectorizer để chuyển dữ liệu đầu vào từ định dạng Text thành Vectơ**","metadata":{}},{"cell_type":"code","source":"from sklearn.linear_model import LogisticRegression\nfrom sklearn.pipeline import Pipeline\nfrom sklearn.metrics import accuracy_score, f1_score\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.feature_extraction.text import CountVectorizer\n\ncount_vectorizer = CountVectorizer()\nmodel = LogisticRegression(C=1, random_state=0)\n\nvectorize_model_pipeline = Pipeline([\n    ('count_vectorizer', count_vectorizer),\n    ('model', model)\n])","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:54:16.035317Z","iopub.execute_input":"2022-01-07T15:54:16.035768Z","iopub.status.idle":"2022-01-07T15:54:16.049305Z","shell.execute_reply.started":"2022-01-07T15:54:16.035735Z","shell.execute_reply":"2022-01-07T15:54:16.048197Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_train, X_test, y_train, y_test = train_test_split(df_train['question_text_cleaned'], df_train['target'], test_size = 0.3)\nvectorize_model_pipeline.fit(X_train, y_train)","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:54:16.051081Z","iopub.execute_input":"2022-01-07T15:54:16.051845Z","iopub.status.idle":"2022-01-07T15:54:55.587056Z","shell.execute_reply.started":"2022-01-07T15:54:16.051797Z","shell.execute_reply":"2022-01-07T15:54:55.585625Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"predictions = vectorize_model_pipeline.predict(X_test)","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:54:55.588829Z","iopub.execute_input":"2022-01-07T15:54:55.58929Z","iopub.status.idle":"2022-01-07T15:55:00.77546Z","shell.execute_reply.started":"2022-01-07T15:54:55.589244Z","shell.execute_reply":"2022-01-07T15:55:00.774477Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from sklearn.metrics import classification_report\n\nprint(classification_report(y_test, predictions))","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:55:00.858125Z","iopub.execute_input":"2022-01-07T15:55:00.858657Z","iopub.status.idle":"2022-01-07T15:55:01.578137Z","shell.execute_reply.started":"2022-01-07T15:55:00.858623Z","shell.execute_reply":"2022-01-07T15:55:01.577131Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"**Tương tự ta sẽ tiến hành với tập test**","metadata":{}},{"cell_type":"code","source":"# Clean trong tập test\ndf_test['question_text_cleaned'] = df_test.question_text.apply(lambda x: clean(x, True))","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:55:01.579668Z","iopub.execute_input":"2022-01-07T15:55:01.580317Z","iopub.status.idle":"2022-01-07T15:57:40.726555Z","shell.execute_reply.started":"2022-01-07T15:55:01.580266Z","shell.execute_reply":"2022-01-07T15:57:40.725633Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_test['prediction'] = vectorize_model_pipeline.predict(df_test['question_text_cleaned'])","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:57:40.727631Z","iopub.execute_input":"2022-01-07T15:57:40.728096Z","iopub.status.idle":"2022-01-07T15:57:45.720674Z","shell.execute_reply.started":"2022-01-07T15:57:40.728065Z","shell.execute_reply":"2022-01-07T15:57:45.71992Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_final = df_test[['qid','prediction']]\ndf_final.set_index('qid', inplace = True)","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:57:45.721663Z","iopub.execute_input":"2022-01-07T15:57:45.722099Z","iopub.status.idle":"2022-01-07T15:57:45.81277Z","shell.execute_reply.started":"2022-01-07T15:57:45.722069Z","shell.execute_reply":"2022-01-07T15:57:45.81179Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_final.head(10)","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:57:45.814048Z","iopub.execute_input":"2022-01-07T15:57:45.814334Z","iopub.status.idle":"2022-01-07T15:57:45.823383Z","shell.execute_reply.started":"2022-01-07T15:57:45.814307Z","shell.execute_reply":"2022-01-07T15:57:45.822594Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Ghi lại kết quả\ndf_final.to_csv('submission.csv')","metadata":{"execution":{"iopub.status.busy":"2022-01-07T15:57:45.825998Z","iopub.execute_input":"2022-01-07T15:57:45.826404Z","iopub.status.idle":"2022-01-07T15:57:46.380419Z","shell.execute_reply.started":"2022-01-07T15:57:45.826372Z","shell.execute_reply":"2022-01-07T15:57:46.37909Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 3. Đánh giá, kết luận\n**Kết quả khi thực hiện mô hình Logistic Regression:**\n\nRun: **545.6s**\n\nPrivate Score: **0.52156**\n\nPublic Score: **0.51618**\n\nBest Score: **0.52156 V1**\n","metadata":{}}]}