{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[],"authorship_tag":"ABX9TyNOJwDMyBGlz7c7ob72JGAm"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"}},"cells":[{"cell_type":"code","execution_count":null,"metadata":{"id":"zDSFdp5iA8J8"},"outputs":[],"source":["import re"]},{"cell_type":"code","source":["# Sample documents\n","docs = {\n","    1: \"Information Retrieval has 2 models and information\",\n","    2: \"Boolean is a basic Information Retrieval classic model\",\n","    3: \"Information is a data that processed, Information\",\n","    4: \"When a Data Processed the result is Information, Data.\"\n","}"],"metadata":{"id":"oYRq_ixmpROc"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# All doc IDs\n","all_docs = set(docs.keys())"],"metadata":{"id":"uS8-HXCdpSdE"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Search terms\n","terms = [\"information\", \"data\", \"retrieval\"]"],"metadata":{"id":"LnCtn3NIpYHH"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Function to check if term exists in a doc\n","def has_term(doc, term):\n","    words = set(re.findall(r'\\w+', doc.lower()))\n","    return term in words"],"metadata":{"id":"ir-tVZWCpZP3"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Collect docs containing each term\n","data_docs = set(i for i, doc in docs.items() if has_term(doc, \"data\"))\n","info_docs = set(i for i, doc in docs.items() if has_term(doc, \"information\"))\n","retrieval_docs = set(i for i, doc in docs.items() if has_term(doc, \"retrieval\"))"],"metadata":{"id":"MA5WehdQpbSq"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Evaluate query: (Data AND Information) OR (NOT Retrieval)\n","result = (data_docs & info_docs) | (all_docs - retrieval_docs)"],"metadata":{"id":"GOFfQZp4pd1p"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["# Output results\n","print(\"Result for Query : (Data ^ Information) v (~ Retrieval)\\n\")\n","for i in sorted(result):\n","    print(f\"Doc{i}:\", docs[i])"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"A175DGb3pggE","executionInfo":{"status":"ok","timestamp":1758816773499,"user_tz":-330,"elapsed":56,"user":{"displayName":"YOGESH SAHU","userId":"09036209155327710676"}},"outputId":"489578a9-56a3-4222-b85d-336bef0ca0e1"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Result for Query : (Data ^ Information) v (~ Retrieval)\n","\n","Doc3: Information is a data that processed, Information\n","Doc4: When a Data Processed the result is Information, Data.\n"]}]},{"cell_type":"code","source":["import re\n","\n","docs = {\n","    1: \"Information Retrieval has 2 models and information.\",\n","    2: \"Boolean is a basic Information Retrieval classic model.\",\n","    3: \"Information is a data that processed, Information.\",\n","    4: \"When a Data Processed the result is Information, Data.\"\n","}\n","\n","# All document IDs\n","all_docs = set(docs.keys())\n","\n","# Terms we are interested in\n","terms = [\"information\", \"data\", \"retrieval\"]\n","\n","# Function to check if a term exists in a document\n","def has_term(doc, term):\n","    words = set(re.findall(r'\\w+', doc.lower()))\n","    return term in words\n","\n","# Collect documents containing each term\n","data_docs = set(i for i, doc in docs.items() if has_term(doc, \"data\"))\n","info_docs = set(i for i, doc in docs.items() if has_term(doc, \"information\"))\n","retrieval_docs = set(i for i, doc in docs.items() if has_term(doc, \"retrieval\"))\n","\n","# Evaluate query: (Data AND Information) OR (NOT Retrieval)\n","result = (data_docs & info_docs) | (all_docs - retrieval_docs)\n","\n","print(\"Result for Query: (Data ^ Information) v (~ Retrieval)\\n\")\n","for i in sorted(result):\n","    print(f\"Doc{i}:\", docs[i])\n"],"metadata":{"id":"-5ZYAh9sMIVj","executionInfo":{"status":"ok","timestamp":1759597609309,"user_tz":-330,"elapsed":31,"user":{"displayName":"YOGESH SAHU","userId":"09036209155327710676"}},"outputId":"d2be560d-fcf1-4328-a3ae-7b3ea11c3703","colab":{"base_uri":"https://localhost:8080/"}},"execution_count":1,"outputs":[{"output_type":"stream","name":"stdout","text":["Result for Query: (Data ^ Information) v (~ Retrieval)\n","\n","Doc3: Information is a data that processed, Information.\n","Doc4: When a Data Processed the result is Information, Data.\n"]}]}]}