RemotelyBest/RemotelyBest_Development_of_AI_Applications
0
1{2 "cells": [3 {4 "cell_type": "code",5 "execution_count": 14,6 "metadata": {},7 "outputs": [8 {9 "name": "stdout",10 "output_type": "stream",11 "text": [12 "* Running on local URL: http://127.0.0.1:7870\n",13 "* Running on public URL: https://a94e18f722148a0463.gradio.live\n",14 "\n",15 "This share link expires in 72 hours. For free permanent hosting and GPU upgrades, run `gradio deploy` from the terminal in the working directory to deploy to Hugging Face Spaces (https://huggingface.co/spaces)\n"16 ]17 },18 {19 "data": {20 "text/html": [21 "<div><iframe src=\"https://a94e18f722148a0463.gradio.live\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"22 ],23 "text/plain": [24 "<IPython.core.display.HTML object>"25 ]26 },27 "metadata": {},28 "output_type": "display_data"29 },30 {31 "data": {32 "text/plain": []33 },34 "execution_count": 14,35 "metadata": {},36 "output_type": "execute_result"37 }38 ],39 "source": [40 "from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, AutoModelForSequenceClassification, TextClassificationPipeline\n",41 "import torch\n",42 "import gradio as gr\n",43 "from openpyxl import load_workbook\n",44 "from numpy import mean\n",45 "import pandas as pd\n",46 "import matplotlib.pyplot as plt\n",47 "\n",48 "theme = gr.themes.Soft(\n",49 " primary_hue=\"amber\",\n",50 " secondary_hue=\"amber\",\n",51 " neutral_hue=\"stone\",\n",52 ")\n",53 "\n",54 "# Load tokenizers and models\n",55 "tokenizer = AutoTokenizer.from_pretrained(\"suriya7/bart-finetuned-text-summarization\")\n",56 "model = AutoModelForSeq2SeqLM.from_pretrained(\"suriya7/bart-finetuned-text-summarization\")\n",57 "\n",58 "tokenizer_keywords = AutoTokenizer.from_pretrained(\"transformer3/H2-keywordextractor\")\n",59 "model_keywords = AutoModelForSeq2SeqLM.from_pretrained(\"transformer3/H2-keywordextractor\")\n",60 "\n",61 "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",62 "new_model = AutoModelForSequenceClassification.from_pretrained('roberta-rating')\n",63 "new_tokenizer = AutoTokenizer.from_pretrained('roberta-rating')\n",64 "\n",65 "classifier = TextClassificationPipeline(model=new_model, tokenizer=new_tokenizer, device=device)\n",66 "\n",67 "label_mapping = {1: '1/5', 2: '2/5', 3: '3/5', 4: '4/5', 5: '5/5'}\n",68 "\n",69 "# Function to display and filter the Excel workbook\n",70 "def filter_xl(file, keywords):\n",71 " # Load the workbook and convert it to a DataFrame\n",72 " workbook = load_workbook(filename=file)\n",73 " sheet = workbook.active\n",74 " data = sheet.values\n",75 " columns = next(data)[0:]\n",76 " df = pd.DataFrame(data, columns=columns)\n",77 " \n",78 " if keywords:\n",79 " keyword_list = keywords.split(',')\n",80 " for keyword in keyword_list:\n",81 " df = df[df.apply(lambda row: row.astype(str).str.contains(keyword.strip(), case=False).any(), axis=1)]\n",82 " \n",83 " return df\n",84 "\n",85 "# Function to calculate overall rating from filtered data\n",86 "def calculate_rating(filtered_df):\n",87 " reviews = filtered_df.to_numpy().flatten()\n",88 " ratings = []\n",89 " for review in reviews:\n",90 " if pd.notna(review):\n",91 " rating = int(classifier(review)[0]['label'].split('_')[1])\n",92 " ratings.append(rating)\n",93 " \n",94 " return round(mean(ratings), 2), ratings\n",95 "\n",96 "# Function to calculate results including summary, keywords, and sentiment\n",97 "def calculate_results(file, keywords):\n",98 " filtered_df = filter_xl(file, keywords)\n",99 " overall_rating, ratings = calculate_rating(filtered_df)\n",100 " \n",101 " # Summarize and extract keywords from the filtered reviews\n",102 " text = \" \".join(filtered_df.to_numpy().flatten())\n",103 " inputs = tokenizer([text], max_length=1024, truncation=True, return_tensors=\"pt\")\n",104 " summary_ids = model.generate(inputs[\"input_ids\"], num_beams=2, min_length=10, max_length=50)\n",105 " summary = tokenizer.batch_decode(summary_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]\n",106 " summary = summary.replace(\"I\", \"They\").replace(\"my\", \"their\").replace(\"me\", \"them\")\n",107 "\n",108 " inputs_keywords = tokenizer_keywords([text], max_length=1024, truncation=True, return_tensors=\"pt\")\n",109 " summary_ids_keywords = model_keywords.generate(inputs_keywords[\"input_ids\"], num_beams=2, min_length=0, max_length=100)\n",110 " keywords = tokenizer_keywords.batch_decode(summary_ids_keywords, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]\n",111 "\n",112 " # Determine overall sentiment\n",113 " sentiments = []\n",114 " for review in filtered_df.to_numpy().flatten():\n",115 " if pd.notna(review):\n",116 " sentiment = classifier(review)[0]['label']\n",117 " sentiment_label = \"Positive\" if sentiment == \"LABEL_4\" or sentiment == \"LABEL_5\" else \"Negative\" if sentiment == \"LABEL_1\" or sentiment == \"LABEL_2\" else \"Neutral\"\n",118 " sentiments.append(sentiment_label)\n",119 " \n",120 " overall_sentiment = \"Positive\" if sentiments.count(\"Positive\") > sentiments.count(\"Negative\") else \"Negative\" if sentiments.count(\"Negative\") > sentiments.count(\"Positive\") else \"Neutral\"\n",121 "\n",122 " return overall_rating, summary, keywords, overall_sentiment, ratings, sentiments\n",123 "\n",124 "# Function to analyze a single review\n",125 "def analyze_review(review):\n",126 " if not review.strip():\n",127 " return \"Error: No text provided\", \"Error: No text provided\", \"Error: No text provided\", \"Error: No text provided\"\n",128 " \n",129 " # Calculate rating\n",130 " rating = int(classifier(review)[0]['label'].split('_')[1])\n",131 " \n",132 " # Summarize review\n",133 " inputs = tokenizer([review], max_length=1024, truncation=True, return_tensors=\"pt\")\n",134 " summary_ids = model.generate(inputs[\"input_ids\"], num_beams=2, min_length=10, max_length=50)\n",135 " summary = tokenizer.batch_decode(summary_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]\n",136 " summary = summary.replace(\"I\", \"he/she\").replace(\"my\", \"his/her\").replace(\"me\", \"him/her\")\n",137 "\n",138 " # Extract keywords\n",139 " inputs_keywords = tokenizer_keywords([review], max_length=1024, truncation=True, return_tensors=\"pt\")\n",140 " summary_ids_keywords = model_keywords.generate(inputs_keywords[\"input_ids\"], num_beams=2, min_length=0, max_length=100)\n",141 " keywords = tokenizer_keywords.batch_decode(summary_ids_keywords, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]\n",142 "\n",143 " # Determine sentiment\n",144 " sentiment = classifier(review)[0]['label']\n",145 " sentiment_label = \"Positive\" if sentiment == \"LABEL_4\" or sentiment == \"LABEL_5\" else \"Negative\" if sentiment == \"LABEL_1\" or sentiment == \"LABEL_2\" else \"Neutral\"\n",146 "\n",147 " return rating, summary, keywords, sentiment_label\n",148 "\n",149 "# Function to count rows in the filtered DataFrame\n",150 "def count_rows(filtered_df):\n",151 " return len(filtered_df)\n",152 "\n",153 "# Function to plot ratings\n",154 "def plot_ratings(ratings):\n",155 " plt.figure(figsize=(10, 5))\n",156 " plt.hist(ratings, bins=range(1, 7), edgecolor='black', align='left')\n",157 " plt.xlabel('Rating')\n",158 " plt.ylabel('Frequency')\n",159 " plt.title('Distribution of Ratings')\n",160 " plt.xticks(range(1, 6))\n",161 " plt.grid(True)\n",162 " plt.savefig('ratings_distribution.png')\n",163 " return 'ratings_distribution.png'\n",164 "\n",165 "# Function to plot sentiments\n",166 "def plot_sentiments(sentiments):\n",167 " sentiment_counts = pd.Series(sentiments).value_counts()\n",168 " plt.figure(figsize=(10, 5))\n",169 " sentiment_counts.plot(kind='bar', color=['green', 'red', 'blue'])\n",170 " plt.xlabel('Sentiment')\n",171 " plt.ylabel('Frequency')\n",172 " plt.title('Distribution of Sentiments')\n",173 " plt.grid(True)\n",174 " plt.savefig('sentiments_distribution.png')\n",175 " return 'sentiments_distribution.png'\n",176 "\n",177 "# Gradio interface\n",178 "with gr.Blocks(theme=theme) as demo:\n",179 " gr.Markdown(\"<h1 style='text-align: center;'>Feedback and Auditing Survey AI Analyzer</h1><br>\")\n",180 " with gr.Tabs():\n",181 " with gr.TabItem(\"Upload and Filter\"):\n",182 " with gr.Row():\n",183 " with gr.Column(scale=1):\n",184 " excel_file = gr.File(label=\"Upload Excel File\")\n",185 " #excel_file = gr.File(label=\"Upload Excel File\", file_types=[\".xlsx\", \".xlsm\", \".xltx\", \".xltm\"])\n",186 " keywords_input = gr.Textbox(label=\"Filter by Keywords (comma-separated)\")\n",187 " display_button = gr.Button(\"Display and Filter Excel Data\")\n",188 " clear_button_upload = gr.Button(\"Clear\")\n",189 " row_count = gr.Textbox(label=\"Number of Rows\", interactive=False)\n",190 " with gr.Column(scale=3):\n",191 " filtered_data = gr.Dataframe(label=\"Filtered Excel Contents\")\n",192 " \n",193 " with gr.TabItem(\"Calculate Results\"):\n",194 " with gr.Row():\n",195 " with gr.Column():\n",196 " overall_rating = gr.Textbox(label=\"Overall Rating\")\n",197 " summary = gr.Textbox(label=\"Summary\")\n",198 " keywords_output = gr.Textbox(label=\"Keywords\")\n",199 " overall_sentiment = gr.Textbox(label=\"Overall Sentiment\")\n",200 " calculate_button = gr.Button(\"Calculate Results\")\n",201 " with gr.Column():\n",202 " ratings_graph = gr.Image(label=\"Ratings Distribution\")\n",203 " sentiments_graph = gr.Image(label=\"Sentiments Distribution\")\n",204 " calculate_graph_button = gr.Button(\"Calculate Graph Results\")\n",205 " \n",206 " with gr.TabItem(\"Testing Area / Write a Review\"):\n",207 " with gr.Row():\n",208 " with gr.Column(scale=2):\n",209 " review_input = gr.Textbox(label=\"Write your review here\")\n",210 " analyze_button = gr.Button(\"Analyze Review\")\n",211 " clear_button_review = gr.Button(\"Clear\")\n",212 " with gr.Column(scale=2):\n",213 " review_rating = gr.Textbox(label=\"Rating\")\n",214 " review_summary = gr.Textbox(label=\"Summary\")\n",215 " review_keywords = gr.Textbox(label=\"Keywords\")\n",216 " review_sentiment = gr.Textbox(label=\"Sentiment\")\n",217 "\n",218 " display_button.click(lambda file, keywords: (filter_xl(file, keywords), count_rows(filter_xl(file, keywords))), inputs=[excel_file, keywords_input], outputs=[filtered_data, row_count])\n",219 " calculate_graph_button.click(lambda file, keywords: (*calculate_results(file, keywords)[:4], plot_ratings(calculate_results(file, keywords)[4]), plot_sentiments(calculate_results(file, keywords)[5])), inputs=[excel_file, keywords_input], outputs=[overall_rating, summary, keywords_output, overall_sentiment, ratings_graph, sentiments_graph])\n",220 " calculate_button.click(lambda file, keywords: (*calculate_results(file, keywords)[:4], plot_ratings(calculate_results(file, keywords)[4])), inputs=[excel_file, keywords_input], outputs=[overall_rating, summary, keywords_output, overall_sentiment])\n",221 " analyze_button.click(analyze_review, inputs=review_input, outputs=[review_rating, review_summary, review_keywords, review_sentiment])\n",222 " clear_button_upload.click(lambda: (\"\"), outputs=[keywords_input])\n",223 " clear_button_review.click(lambda: (\"\", \"\", \"\", \"\", \"\"), outputs=[review_input, review_rating, review_summary, review_keywords, review_sentiment])\n",224 "\n",225 "demo.launch(share=True)"226 ]227 }228 ],229 "metadata": {230 "kernelspec": {231 "display_name": "SolutionsInPR",232 "language": "python",233 "name": "python3"234 },235 "language_info": {236 "codemirror_mode": {237 "name": "ipython",238 "version": 3239 },240 "file_extension": ".py",241 "mimetype": "text/x-python",242 "name": "python",243 "nbconvert_exporter": "python",244 "pygments_lexer": "ipython3",245 "version": "3.12.4"246 }247 },248 "nbformat": 4,249 "nbformat_minor": 2250}251