azure-ai-search-maximum-marginal-relevance.ipynb

{
 "cells": [
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Enhancing Answer Quality in Retrieval-Augmented Generation (RAG) with Maximum Marginal Relevance (MMR)\n",
    "\n",
    "This notebook demonstrates how integrating Maximum Marginal Relevance (MMR) into a Retrieval-Augmented Generation (RAG) pipeline can improve the quality of synthesized answers. While relevance is crucial, MMR goes a step further by ensuring the selected documents provide diverse perspectives. This can lead to more comprehensive, well-rounded answers during the final synthesis stage (L3) of RAG.\n",
    "\n",
    "**Why MMR in a RAG Context?**  \n",
    "Retrieval-Augmented Generation relies on retrieved documents (context) to guide the large language model (LLM) in crafting accurate and insightful responses. However, if the top documents are too similar—only echoing the same points—your model may miss critical nuances. MMR helps mitigate this by balancing relevance with diversity, ensuring that the retrieved set of documents covers different angles and subtopics.\n",
    "\n",
    "**Key Concepts:**\n",
    "\n",
    "- **Relevance:** How well a document matches the user's query.\n",
    "- **Diversity:** Ensuring the returned documents are not overly similar to each other, thus offering a broader coverage of relevant information.\n",
    "- **Trade-off Parameter (λ):** Controls the balance between relevance and diversity (0 ≤ λ ≤ 1)\n",
    "  - λ = 1: Pure relevance ranking\n",
    "  - λ = 0: Maximum diversity\n",
    "  - λ = 0.5: Equal weight to both\n",
    "\n",
    "**MMR Formula:**\n",
    "\n",
    "![image.png](attachment:image.png)\n",
    "\n",
    "\n",
    "This implementation provides a foundation for more sophisticated search result diversification strategies while maintaining relevance to the original query."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install azure-search-documents==11.6.0b8\n",
    "!pip install openai==1.45.0\n",
    "!pip install python-dotenv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "from typing import List, Dict, Any\n",
    "from azure.search.documents import SearchClient\n",
    "from azure.core.credentials import AzureKeyCredential\n",
    "from azure.identity import DefaultAzureCredential\n",
    "from azure.search.documents.indexes import SearchIndexClient\n",
    "from azure.search.documents.models import (\n",
    "    VectorizableTextQuery,\n",
    "    VectorizedQuery\n",
    ")\n",
    "from openai import AzureOpenAI\n",
    "from dotenv import load_dotenv\n",
    "\n",
    "load_dotenv()\n",
    "\n",
    "# Environment variables\n",
    "AZURE_SEARCH_SERVICE_ENDPOINT = os.getenv(\"AZURE_SEARCH_SERVICE_ENDPOINT\")\n",
    "AZURE_SEARCH_ADMIN_KEY = os.getenv(\"AZURE_SEARCH_ADMIN_KEY\")\n",
    "AZURE_OPENAI_ENDPOINT = os.getenv(\"AZURE_OPENAI_ENDPOINT\")\n",
    "AZURE_OPENAI_API_KEY = os.getenv(\"AZURE_OPENAI_API_KEY\")\n",
    "AZURE_OPENAI_API_VERSION = \"2024-10-01-preview\"\n",
    "AZURE_OPENAI_EMBEDDING_DEPLOYED_MODEL_NAME = \"text-embedding-3-large\"\n",
    "\n",
    "credential = AzureKeyCredential(AZURE_SEARCH_ADMIN_KEY)\n",
    "index_name = \"compression-test-baseline\"\n",
    "search_client = SearchClient(endpoint=AZURE_SEARCH_SERVICE_ENDPOINT, index_name=index_name, credential=credential)\n",
    "\n",
    "# Configure AzureOpenAI client (API Key auth)\n",
    "client = AzureOpenAI(\n",
    "    api_key=AZURE_OPENAI_API_KEY,\n",
    "    api_version=AZURE_OPENAI_API_VERSION,\n",
    "    azure_endpoint=AZURE_OPENAI_ENDPOINT\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Environment variables\n",
    "AZURE_SEARCH_SERVICE_ENDPOINT = os.getenv(\"AZURE_SEARCH_SERVICE_ENDPOINT\")\n",
    "AZURE_SEARCH_ADMIN_KEY = os.getenv(\"AZURE_SEARCH_ADMIN_KEY\")\n",
    "AZURE_OPENAI_ENDPOINT: str = os.getenv(\"AZURE_OPENAI_ENDPOINT\")\n",
    "AZURE_OPENAI_API_KEY: str = os.getenv(\"AZURE_OPENAI_API_KEY\")\n",
    "AZURE_OPENAI_API_VERSION: str = \"2024-10-01-preview\"\n",
    "AZURE_OPENAI_EMBEDDING_DEPLOYED_MODEL_NAME: str = \"text-embedding-3-large\"\n",
    "\n",
    "credential = AzureKeyCredential(AZURE_SEARCH_ADMIN_KEY)\n",
    "index_name = \"compression-test-baseline\"\n",
    "search_client = SearchClient(endpoint=AZURE_SEARCH_SERVICE_ENDPOINT, index_name=index_name, credential=credential)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_embeddings(text: str):\n",
    "    # Generate embeddings using Azure OpenAI\n",
    "    embeddings_response = client.embeddings.create(\n",
    "        model=AZURE_OPENAI_EMBEDDING_DEPLOYED_MODEL_NAME, \n",
    "        input=text, \n",
    "        dimensions=3072\n",
    "    )\n",
    "    return embeddings_response.data[0].embedding\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Query Embedding:\n",
      "[0.024452010169625282, 0.011633044108748436, -0.001606191392056644, -0.02325386181473732, -0.028926728293299675]...\n",
      "--------------------------------------------------------------------------------\n",
      "Number of results: 50\n",
      "title: Blairism\n",
      "content: In British politics, the term Blairism refers to the political ideology of former leader of the Labour Party and Prime Minister Tony Blair, who left both positions in 2007 to become Envoy of the Quart...\n",
      "embedding: [0.0010587202, 0.023848193, -0.015566647, -0.030624002, -0.05619936]...\n",
      "@search.score: 0.6392048\n",
      "--------------------------------------------------------------------------------\n",
      "title: Jonathan Powell (Labour adviser)\n",
      "content: Jonathan Nicholas Powell (born 14 August 1956) is a British diplomat who served as the first Downing Street Chief of Staff, under British Prime Minister Tony Blair from 1995 to 2007. He was the only s...\n",
      "embedding: [0.01429064, 0.002617065, -0.011580346, -0.021988673, -0.039235998]...\n",
      "@search.score: 0.6245392\n",
      "--------------------------------------------------------------------------------\n",
      "title: Vicky Pope\n",
      "content: Vicky Pope is head of the climate predictions programme at the Hadley Centre. She spent 6 years as manager of atmospheric climate model development and evaluation. Since October 2004 she has been Head...\n",
      "embedding: [0.010052227, 0.01045994, -0.016674044, 0.017897183, 0.018403308]...\n",
      "@search.score: 0.6153679\n",
      "--------------------------------------------------------------------------------\n",
      "title: Tony Abbott\n",
      "content: Anthony John \"Tony\" Abbott (born 4 November 1957) is an Australian politician who was the 28th Prime Minister of Australia from 18 September 2013 to 15 September 2015. Abbott was leader of the Liberal...\n",
      "embedding: [0.001317827, -0.0025967753, 0.008520923, 0.008365408, -0.010970282]...\n",
      "@search.score: 0.60499316\n",
      "--------------------------------------------------------------------------------\n",
      "title: Earth in the Balance\n",
      "content: Earth in the Balance: Ecology and the Human Spirit (ISBN 0-452-26935-0, paperback ISBN 1-85383-743-1) is a 1992 book written by Al Gore, published in June 1992, shortly before he was elected Vice Pres...\n",
      "embedding: [-0.0021811717, -0.03816736, -0.0150356265, -0.014733909, -0.044503443]...\n",
      "@search.score: 0.6040373\n",
      "--------------------------------------------------------------------------------\n",
      "title: Hadley Centre for Climate Prediction and Research\n",
      "content: The Met Office Hadley Centre for Climate Change — named in honour of George Hadley —  is one of the United Kingdom's leading centres for the study of scientific issues associated with climate change. ...\n",
      "embedding: [-0.023889415, 0.005361545, -0.009121413, 0.01149678, 0.0032508606]...\n",
      "@search.score: 0.59642917\n",
      "--------------------------------------------------------------------------------\n",
      "title: Jack Straw\n",
      "content: John Whitaker Straw (born 3 August 1946) is a British politician who served as the Member of Parliament (MP) for Blackburn from 1979 to 2015. Straw served in the Cabinet from 1997 to 2010 under the go...\n",
      "embedding: [0.019625338, -0.012640251, -0.0054994076, -0.011520001, -0.040281065]...\n",
      "@search.score: 0.5961766\n",
      "--------------------------------------------------------------------------------\n",
      "title: Minister of Climate and the Environment\n",
      "content: The Minister of Climate and the Environment (Norwegian: Klima- og miljøministeren) is a Councilor of State and Chief of Norway's Ministry of the Environment. Since 16 October 2013 the position has bee...\n",
      "embedding: [-0.002048663, -0.008697125, -0.008448991, -0.035235144, -0.006612793]...\n",
      "@search.score: 0.59415525\n",
      "--------------------------------------------------------------------------------\n",
      "title: John Gummer\n",
      "content: John Selwyn Gummer, Baron Deben, PC (born 26 November 1939 at Stockport, Cheshire) is a British Conservative Party politician, formerly Member of Parliament (MP) for Suffolk Coastal and now a member o...\n",
      "embedding: [0.025042715, 0.014129605, -0.016975949, -0.00751792, 0.011940601]...\n",
      "@search.score: 0.5929634\n",
      "--------------------------------------------------------------------------------\n",
      "title: Abrupt climate change\n",
      "content: An abrupt climate change occurs when the climate system is forced to transition to a new climate state at a rate that is determined by the climate system energy-balance, and which is more rapid than t...\n",
      "embedding: [-0.012501736, -0.016382972, -0.011523047, -0.008553467, -0.05338543]...\n",
      "@search.score: 0.58912444\n",
      "--------------------------------------------------------------------------------\n",
      "title: Frank Sartor\n",
      "content: Francesco Ernest \"Frank\" Sartor AO (born 9 November 1951) a former Australian politician, was New South Wales Minister for Climate Change and the Environment and Minister Assisting the Minister for He...\n",
      "embedding: [0.007416533, -0.021494461, -0.022815952, -0.0413438, 0.0056702765]...\n",
      "@search.score: 0.58906037\n",
      "--------------------------------------------------------------------------------\n",
      "title: Shaun Woodward\n",
      "content: Shaun Anthony Woodward (born 26 October 1958) is a British politician who was the Labour Member of Parliament (MP) for St Helens South from 2001 to 2015. He served in the cabinet from 28 June 2007 to ...\n",
      "embedding: [-0.015324105, 0.03550707, -0.007889811, -0.02192328, -0.024107432]...\n",
      "@search.score: 0.58890504\n",
      "--------------------------------------------------------------------------------\n",
      "title: United Nations Framework Convention on Climate Change\n",
      "content: The United Nations Framework Convention on Climate Change (UNFCCC) is an international environmental treaty (currently the only international climate policy venue with broad legitimacy, due in part to...\n",
      "embedding: [0.005576988, -0.008540075, -0.015350851, -0.028094452, 0.0015788164]...\n",
      "@search.score: 0.5882966\n",
      "--------------------------------------------------------------------------------\n",
      "title: George Brown, Baron George-Brown\n",
      "content: George Alfred Brown, Baron George-Brown, PC (2 September 1914 – 2 June 1985) was a British Labour politician who served as Deputy Leader of the Labour Party from 1960 to 1970 and also in a number of C...\n",
      "embedding: [0.03435035, 0.0007764186, -0.013161689, -0.038223684, -0.007956897]...\n",
      "@search.score: 0.5876888\n",
      "--------------------------------------------------------------------------------\n",
      "title: Charles Leadbeater\n",
      "content: Charles Leadbeater (formerly known as Charlie Leadbeater) is a British author and former advisor to Tony Blair.A graduate of Balliol College, Oxford, he first came to widespread notice in the 1980s as...\n",
      "embedding: [0.011149497, -0.007662695, -0.022781357, -0.020782992, -0.065877125]...\n",
      "@search.score: 0.5866875\n",
      "--------------------------------------------------------------------------------\n",
      "title: Timetable For Action\n",
      "content: The Timetable For Action is a policy document by the U.K. Conservative Party, in which leader Michael Howard spells out some key actions he would take if elected Prime Minister, within the first hour,...\n",
      "embedding: [0.00575742, -0.01976843, -0.010309259, -0.0020614655, -0.014745177]...\n",
      "@search.score: 0.5852342\n",
      "--------------------------------------------------------------------------------\n",
      "title: Bert Bolin\n",
      "content: Bert Rickard Johannes Bolin (Swedish: [bæʈː bʊliːn]; 15 May 1925 – 30 December 2007) was a Swedish meteorologist who served as the first chairman of the Intergovernmental Panel on Climate Change (IPCC...\n",
      "embedding: [0.0007309993, -0.037017774, -0.016030349, -0.0074602775, -0.0041247318]...\n",
      "@search.score: 0.5851088\n",
      "--------------------------------------------------------------------------------\n",
      "title: James Croll\n",
      "content: James Croll, FRS,  (2 January 1821 – 15 December 1890) was a 19th-century Scottish scientist who developed a theory of climate change based on changes in the Earth's orbit....\n",
      "embedding: [-0.012882465, -0.020949584, -0.013052418, 0.001033883, -0.035033055]...\n",
      "@search.score: 0.58457595\n",
      "--------------------------------------------------------------------------------\n",
      "title: Jim McLay\n",
      "content: Sir James Kenneth \"Jim\" McLay KNZM QSO (born 21 February 1945) is a New Zealand diplomat and former politician. He was Deputy Prime Minister, leader of the National Party and Leader of the Opposition ...\n",
      "embedding: [0.03415941, -0.008323815, -0.021400465, -0.013368936, -0.06298141]...\n",
      "@search.score: 0.5845506\n",
      "--------------------------------------------------------------------------------\n",
      "title: William Connolley\n",
      "content: William Michael Connolley (born 12 April 1964) is a British software engineer, writer, and blogger on climatology. Until December 2007 he was Senior Scientific Officer in the Physical Sciences Divisio...\n",
      "embedding: [0.010537665, -0.0045892093, -0.02509702, -0.03752642, -0.024088094]...\n",
      "@search.score: 0.5845407\n",
      "--------------------------------------------------------------------------------\n",
      "title: Tony Cunningham\n",
      "content: Sir Thomas Anthony Cunningham (born 16 September 1952), known as Tony Cunningham, is a British Labour Party  politician who was the Member of Parliament (MP) for Workington from 2001 to 2015. Ed Milib...\n",
      "embedding: [0.03471322, 0.00414721, -0.016689532, -0.020251475, -0.018980252]...\n",
      "@search.score: 0.583494\n",
      "--------------------------------------------------------------------------------\n",
      "title: Global warming controversy\n",
      "content: The global warming controversy concerns the public debate over whether global warming is occurring, how much has occurred in modern times, what has caused it, what its effects will be, whether any act...\n",
      "embedding: [-0.010281988, -0.016738687, -0.016080836, -0.0044618226, -0.02406034]...\n",
      "@search.score: 0.58317894\n",
      "--------------------------------------------------------------------------------\n",
      "title: Prime Minister's Delivery Unit\n",
      "content: The Prime Minister’s Delivery Unit was established in June 2001 to monitor progress on and strengthen the British Government’s capacity to deliver its key priorities across education, health, crime an...\n",
      "embedding: [0.016872521, 0.04073826, -0.007312351, 0.014485948, 0.003593736]...\n",
      "@search.score: 0.5821886\n",
      "--------------------------------------------------------------------------------\n",
      "title: Clive Ponting\n",
      "content: Clive Sheridan Ponting (born 13 April 1946) is a former senior civil servant, best known for leaking documents about the sinking of the Belgrano in the Falklands War. He is the author of a number of r...\n",
      "embedding: [0.0055358484, -0.004278756, -0.021188136, -0.013161322, -0.02352321]...\n",
      "@search.score: 0.5808566\n",
      "--------------------------------------------------------------------------------\n",
      "title: S. Robert Blair\n",
      "content: Sidney Robert \"Bob\" Blair, CC (13 August 1929 – 18 April 2009) was a Canadian engineer and businessman.Born in Port of Spain, Trinidad, where his father Sidney Blair was managing an oil refinery, the ...\n",
      "embedding: [0.021150235, -0.010680606, -0.013258455, -0.027057534, 0.0071006245]...\n",
      "@search.score: 0.58030486\n",
      "--------------------------------------------------------------------------------\n",
      "title: Anthony Lake\n",
      "content: William Anthony Kirsopp Lake, best known as Tony Lake, (born April 2, 1939) is the Executive Director of the United Nations Children's Fund (UNICEF), author, academic, and former American diplomat, Fo...\n",
      "embedding: [0.00440076, -0.008354777, -0.015602695, -0.004984194, -0.04440767]...\n",
      "@search.score: 0.57980096\n",
      "--------------------------------------------------------------------------------\n",
      "title: Bertie Ahern\n",
      "content: Patrick Bartholomew \"Bertie\" Ahern (born 12 September 1951) is a former Irish Fianna Fáil politician who served as Taoiseach of Ireland from 26 June 1997 to 7 May 2008.Ahern served as a Teachta Dála (...\n",
      "embedding: [0.0082776835, 0.0013741314, -0.013206015, -0.03423516, -0.011052861]...\n",
      "@search.score: 0.5794539\n",
      "--------------------------------------------------------------------------------\n",
      "title: Singaporean response to Hurricane Katrina\n",
      "content: Following that devastation of the United States Gulf Coast by Hurricane Katrina, the Prime Minister of Singapore, Lee Hsien Loong, sent a letter to US President George W. Bush promising support for th...\n",
      "embedding: [0.029475367, 0.019027762, -0.008024972, -0.011280385, 0.0016560968]...\n",
      "@search.score: 0.5791076\n",
      "--------------------------------------------------------------------------------\n",
      "title: Lester R. Brown\n",
      "content: Lester Russel Brown (born March 28, 1934) is a United States environmental analyst, founder of the Worldwatch Institute, and founder and president of the Earth Policy Institute, a nonprofit research o...\n",
      "embedding: [0.009260519, 0.0015100442, -0.017648388, -0.039073918, -0.024811938]...\n",
      "@search.score: 0.5786437\n",
      "--------------------------------------------------------------------------------\n",
      "title: It's Time\n",
      "content: It's Time was a successful political campaign run by the Australian Labor Party (ALP) under Gough Whitlam at the 1972 election in Australia. Campaigning on the perceived need for change after 23 years...\n",
      "embedding: [-0.0034090523, -0.046031896, -0.014207027, 0.014023801, -0.07340297]...\n",
      "@search.score: 0.5779466\n",
      "--------------------------------------------------------------------------------\n",
      "title: The Prime Minister (novel)\n",
      "content: The Prime Minister is a novel by Anthony Trollope, first published in 1876.  It is the fifth of the \"Palliser\" series of novels....\n",
      "embedding: [-0.03304218, 0.009665286, -0.01583831, -0.040317733, -0.009497389]...\n",
      "@search.score: 0.57780933\n",
      "--------------------------------------------------------------------------------\n",
      "title: Jeremy Browne\n",
      "content: Jeremy Richard Browne (born 17 May 1970) is a British Liberal Democrat politician who was  the Member of Parliament (MP) for Taunton Deane from 2005 to the dissolution of the 55th Parliament on 30 Mar...\n",
      "embedding: [-0.0022932393, -0.006747722, -0.0048591723, -0.016408047, -0.006939188]...\n",
      "@search.score: 0.5775603\n",
      "--------------------------------------------------------------------------------\n",
      "title: Peace for our time\n",
      "content: The phrase \"Peace for Our Time\" was spoken on 30 September 1938 by British Prime Minister Neville Chamberlain in his speech concerning the Munich Agreement and the Anglo-German Declaration....\n",
      "embedding: [0.00062294013, -0.0030225024, -0.015072425, 0.023307743, -0.030529678]...\n",
      "@search.score: 0.5772753\n",
      "--------------------------------------------------------------------------------\n",
      "title: John Howard\n",
      "content: John Winston Howard, OM, AC, (born 26 July 1939) was the 25th Prime Minister of Australia, from 11 March 1996 to 3 December 2007. He is the second longest-serving Australian prime minister, after Sir ...\n",
      "embedding: [0.0390086, 0.006935714, 0.0002833999, -0.017115757, -0.010671803]...\n",
      "@search.score: 0.5771481\n",
      "--------------------------------------------------------------------------------\n",
      "title: Chris Nineham\n",
      "content: Chris Nineham is a founder members and National Officer of the Stop the War Coalition in the UK. He was one of the main organisers of the 15 February 2003 anti-war protest against the invasion on Iraq...\n",
      "embedding: [-0.0029779605, -0.0011286755, -0.028180856, -0.0018903429, -0.03552772]...\n",
      "@search.score: 0.57675153\n",
      "--------------------------------------------------------------------------------\n",
      "title: Tony Ryall\n",
      "content: Anthony Boyd Williams \"Tony\" Ryall CNZM (born 19 November 1964) is a New Zealand politician. He represented the National Party in the New Zealand Parliament from 1990 to 2014. Since November 2008, he ...\n",
      "embedding: [0.018661711, 0.00618773, -0.009642308, -0.038111586, -0.030865554]...\n",
      "@search.score: 0.57650864\n",
      "--------------------------------------------------------------------------------\n",
      "title: Tony Ianno\n",
      "content: Anthony \"Tony\" Ianno, PC (born January 2, 1957) is a businessman and a former Canadian politician.  He served as a Liberal Party of Canada MP representing Trinity—Spadina (1993–2006) and Minister of F...\n",
      "embedding: [-0.015675444, 0.01566193, -0.011371453, -0.038891315, -0.017040288]...\n",
      "@search.score: 0.5764641\n",
      "--------------------------------------------------------------------------------\n",
      "title: Richard Wilson, Baron Wilson of Dinton\n",
      "content: Richard Thomas James Wilson, Baron Wilson of Dinton, GCB (born 11 October 1942) is a cross bench member of the British House of Lords....\n",
      "embedding: [0.048932426, 0.004507098, -0.02170893, 0.018228015, 0.009400965]...\n",
      "@search.score: 0.57629424\n",
      "--------------------------------------------------------------------------------\n",
      "title: Greening Earth Society\n",
      "content: The Greening Earth Society, now defunct, was a public relations organization which promoted the idea that there is considerable scientific doubt about the effects of climate change and increased level...\n",
      "embedding: [-0.00041618393, 0.025366116, -0.01791026, -0.005760729, 0.0045518535]...\n",
      "@search.score: 0.576096\n",
      "--------------------------------------------------------------------------------\n",
      "title: Geoffrey Palmer (politician)\n",
      "content: Sir Geoffrey Winston Russell Palmer KCMG AC QC (born 21 April 1942) served as the 33rd Prime Minister of New Zealand from August 1989 until September 1990, leading the Fourth Labour Government. He was...\n",
      "embedding: [0.04731046, 0.023305763, -0.016581556, -0.0058324663, -0.06613342]...\n",
      "@search.score: 0.5759489\n",
      "--------------------------------------------------------------------------------\n",
      "title: Chris Heaton-Harris\n",
      "content: Christopher Heaton-Harris, (born 28 November 1967), is a British Conservative Party politician. He has been the Member of Parliament (MP) for Daventry since May 2010, and was previously a Member of th...\n",
      "embedding: [-0.017243512, -0.0058539086, -0.024808563, -0.0035213511, 0.0035393632]...\n",
      "@search.score: 0.57565075\n",
      "--------------------------------------------------------------------------------\n",
      "title: Chris Baines\n",
      "content: Chris Baines is an English naturalist, one of the UK's leading independent environmentalists. He is a horticulturalist, landscape architect, naturalist, television presenter and author.Baines grew up ...\n",
      "embedding: [0.021004371, 0.03158069, -0.029331988, -0.0004224041, -0.043491405]...\n",
      "@search.score: 0.5754165\n",
      "--------------------------------------------------------------------------------\n",
      "title: Tony Grounds\n",
      "content: Tony Grounds is a British writer of television scripts. He was born in 1957 in East London.Described by The Independent (11 October 2002) as \"the best TV writer of his generation\", Grounds has written...\n",
      "embedding: [0.032277998, 0.014429835, -0.021490647, -0.01253154, -0.04247695]...\n",
      "@search.score: 0.5750411\n",
      "--------------------------------------------------------------------------------\n",
      "title: Blair toilet\n",
      "content: The Blair Toilet (aka Blair Latrine) is a pit toilet designed in the 1970s. It was a result of large-scale projects to improve rural sanitation in Rhodesia under UDI at the Blair Research Institute, a...\n",
      "embedding: [-0.010243846, 0.0023480032, -0.013258027, -0.043013964, -0.03512176]...\n",
      "@search.score: 0.57491463\n",
      "--------------------------------------------------------------------------------\n",
      "title: David Kelly (weapons expert)\n",
      "content: David Christopher Kelly, CMG (14 May 1944 – 17 July 2003) was a British scientist and authority on biological warfare, employed by the British Ministry of Defence, and formerly a United Nations weapon...\n",
      "embedding: [0.05123183, -0.019895365, -0.021781122, -0.0039265505, -0.031665526]...\n",
      "@search.score: 0.5747966\n",
      "--------------------------------------------------------------------------------\n",
      "title: Tyndall Forum\n",
      "content: The Tyndall Forum is an environmental science and opinion former debating group named after the scientist John Tyndall.It was formed in the UK with the collaboration of the Royal Institution of Great ...\n",
      "embedding: [-0.006833513, -0.028757082, -0.021508519, -0.040437795, 0.034389913]...\n",
      "@search.score: 0.57459295\n",
      "--------------------------------------------------------------------------------\n",
      "title: John Chroston\n",
      "content: John Chroston of Tillicoultry, Clackmannanshire, a biology teacher at Falkirk High School, Scotland, was one of the few tourists present during the Indian Ocean earthquake able to recognise tsunami wa...\n",
      "embedding: [0.03767239, -0.011882581, -0.00856635, 0.0046811225, -0.010954036]...\n",
      "@search.score: 0.5745179\n",
      "--------------------------------------------------------------------------------\n",
      "title: Bernard Ingham\n",
      "content: Sir Bernard Ingham (born 21 June 1932) is a British journalist and former civil servant who is best known as Margaret Thatcher's chief press secretary while she was Prime Minister of the United Kingdo...\n",
      "embedding: [0.013634652, 0.0034056276, -0.025350982, 0.0011298974, -0.0033752746]...\n",
      "@search.score: 0.5743805\n",
      "--------------------------------------------------------------------------------\n",
      "title: Lamberto Dini\n",
      "content: About this sound Lamberto Dini  (born 1 March 1931) is an Italian politician and economist. He was the 51st Prime Minister of Italy from 1995 to 1996 and Foreign Minister from 1996 to 2001....\n",
      "embedding: [-0.033536393, 0.024014754, -0.01137541, -0.023488116, -0.012439221]...\n",
      "@search.score: 0.5742825\n",
      "--------------------------------------------------------------------------------\n",
      "title: World Climate Conference\n",
      "content: The World Climate Conferences are a series of international meetings, organized by the World Meteorological Organization (WMO), about global climate issues principally global warming in addition to cl...\n",
      "embedding: [0.006867968, -0.017415855, -0.0163471, 0.014671097, 0.004609008]...\n",
      "@search.score: 0.5741021\n",
      "--------------------------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "# Pure Vector Search\n",
    "query = \"What did Prime Minister Tony Blair say about climate change?\"  \n",
    "\n",
    "# Generate and print query embedding\n",
    "query_embedding = generate_embeddings(query)\n",
    "print(\"Query Embedding:\")\n",
    "print(f\"{query_embedding[:5]}...\")\n",
    "print(\"-\"*80)\n",
    "\n",
    "# Use a VectorizedQuery with the pre-generated query embedding\n",
    "vector_query = VectorizedQuery(\n",
    "    vector=query_embedding, \n",
    "    k_nearest_neighbors=50, \n",
    "    fields=\"embedding\"\n",
    ")\n",
    "\n",
    "results = search_client.search(  \n",
    "    search_text=None,  \n",
    "    vector_queries=[vector_query],\n",
    "    top=50\n",
    ")  \n",
    "\n",
    "# Convert to a list for multiple iterations\n",
    "results_list = list(results)\n",
    "\n",
    "print(f\"Number of results: {len(results_list)}\")\n",
    "for doc in results_list:\n",
    "    print(f\"title: {doc.get('title', 'No title')}\")\n",
    "    print(f\"content: {doc.get('content', 'No content')[:200]}...\")\n",
    "    print(f\"embedding: {doc.get('embedding', 'No embedding')[:5]}...\")\n",
    "    print(f\"@search.score: {doc.get('@search.score', 'No score')}\")\n",
    "    print(\"-\"*80)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define the MMR Reranking Function\n",
    "\n",
    "Now we define an MMR function that:\n",
    "- Computes relevance as the cosine similarity between the query embedding and document embeddings.\n",
    "- Computes diversity as the maximum cosine similarity between a candidate document and already selected documents.\n",
    "- Uses `lambda_param` to balance between relevance and diversity.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cosine_similarity(vec1: List[float], vec2: List[float]) -> float:\n",
    "    \"\"\"Calculate cosine similarity between two vectors.\"\"\"\n",
    "    vec1_np = np.array(vec1)\n",
    "    vec2_np = np.array(vec2)\n",
    "    denom = (np.linalg.norm(vec1_np) * np.linalg.norm(vec2_np) + 1e-9)\n",
    "    return float(np.dot(vec1_np, vec2_np) / denom)\n",
    "\n",
    "def mmr_reranking(\n",
    "    query_embedding: List[float],\n",
    "    search_results: List[Dict[str, Any]],\n",
    "    lambda_param: float = 0.5,\n",
    "    top_k: int = 10\n",
    ") -> List[Dict[str, Any]]:\n",
    "    \"\"\"\n",
    "    Rerank search results using Maximum Marginal Relevance (MMR) based on embeddings.\n",
    "    \n",
    "    Args:\n",
    "        query_embedding: The embedding vector of the query.\n",
    "        search_results: List of search results, each containing an \"embedding\" field.\n",
    "        lambda_param: Trade-off parameter between relevance (to query) and diversity (0 to 1).\n",
    "        top_k: Number of documents to return.\n",
    "    \n",
    "    Returns:\n",
    "        A reranked list of documents.\n",
    "    \"\"\"\n",
    "    if not search_results:\n",
    "        print(\"Warning: Empty search results provided\")\n",
    "        return []\n",
    "    \n",
    "    selected = []\n",
    "    remaining = search_results.copy()\n",
    "    \n",
    "    print(f\"Starting MMR reranking with {len(remaining)} documents\")\n",
    "    \n",
    "    while len(selected) < top_k and remaining:\n",
    "        mmr_scores = {}\n",
    "        \n",
    "        for i, doc in enumerate(remaining):\n",
    "            # Relevance to the query\n",
    "            relevance_score = cosine_similarity(query_embedding, doc['embedding'])\n",
    "            \n",
    "            # Diversity from already selected docs\n",
    "            if selected:\n",
    "                similarities = [\n",
    "                    cosine_similarity(doc['embedding'], sel_doc['embedding'])\n",
    "                    for sel_doc in selected\n",
    "                ]\n",
    "                diversity_penalty = max(similarities)\n",
    "            else:\n",
    "                diversity_penalty = 0.0\n",
    "            \n",
    "            # MMR Formula\n",
    "            mmr_score = lambda_param * relevance_score - (1 - lambda_param) * diversity_penalty\n",
    "            mmr_scores[i] = mmr_score\n",
    "            \n",
    "            # Debug: Print a sample calculation for the first doc\n",
    "            if len(selected) == 0 and i == 0:\n",
    "                print(\"Sample MMR calculation for first doc:\")\n",
    "                print(f\"  - Relevance score: {relevance_score:.4f}\")\n",
    "                print(f\"  - Diversity penalty: {diversity_penalty:.4f}\")\n",
    "                print(f\"  - Final MMR score: {mmr_score:.4f}\")\n",
    "        \n",
    "        if not mmr_scores:\n",
    "            break\n",
    "        \n",
    "        # Select document with the highest MMR score\n",
    "        next_idx = max(mmr_scores.items(), key=lambda x: x[1])[0]\n",
    "        selected.append(remaining[next_idx])\n",
    "        remaining.pop(next_idx)\n",
    "        \n",
    "        print(f\"Selected document {len(selected)}/{top_k} - Title: {selected[-1]['title']}\")\n",
    "    \n",
    "    return selected\n",
    "\n",
    "def process_vector_search_results(\n",
    "    query_embedding: List[float],\n",
    "    search_results: List[Dict[str, Any]],\n",
    "    lambda_param: float = 0.5,\n",
    "    top_k: int = 10\n",
    ") -> List[Dict[str, Any]]:\n",
    "    \"\"\"\n",
    "    Process search results with MMR reranking using embeddings.\n",
    "    \"\"\"\n",
    "    print(f\"\\nProcessing {len(search_results)} search results\")\n",
    "    print(f\"Lambda parameter: {lambda_param}\")\n",
    "    print(f\"Requested top_k: {top_k}\")\n",
    "    \n",
    "    # Ensure embeddings are NumPy arrays for consistency (if not already)\n",
    "    for doc in search_results:\n",
    "        doc[\"embedding\"] = np.array(doc[\"embedding\"], dtype=float)\n",
    "    \n",
    "    # Apply MMR reranking\n",
    "    reranked_results = mmr_reranking(\n",
    "        query_embedding=query_embedding,\n",
    "        search_results=search_results,\n",
    "        lambda_param=lambda_param,\n",
    "        top_k=min(top_k, len(search_results))\n",
    "    )\n",
    "    \n",
    "    print(f\"\\nReturning {len(reranked_results)} reranked results\")\n",
    "    return reranked_results\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Processing 50 search results\n",
      "Lambda parameter: 0.7\n",
      "Requested top_k: 10\n",
      "Starting MMR reranking with 50 documents\n",
      "Sample MMR calculation for first doc:\n",
      "  - Relevance score: 0.4356\n",
      "  - Diversity penalty: 0.0000\n",
      "  - Final MMR score: 0.3049\n",
      "Selected document 1/10 - Title: Blairism\n",
      "Selected document 2/10 - Title: Vicky Pope\n",
      "Selected document 3/10 - Title: Earth in the Balance\n",
      "Selected document 4/10 - Title: Jonathan Powell (Labour adviser)\n",
      "Selected document 5/10 - Title: Singaporean response to Hurricane Katrina\n",
      "Selected document 6/10 - Title: Abrupt climate change\n",
      "Selected document 7/10 - Title: Frank Sartor\n",
      "Selected document 8/10 - Title: Tony Abbott\n",
      "Selected document 9/10 - Title: United Nations Framework Convention on Climate Change\n",
      "Selected document 10/10 - Title: John Gummer\n",
      "\n",
      "Returning 10 reranked results\n"
     ]
    }
   ],
   "source": [
    "# Apply MMR Reranking\n",
    "reranked_results = process_vector_search_results(\n",
    "    query_embedding=query_embedding,\n",
    "    search_results=results_list,\n",
    "    lambda_param=0.7,  # Bias slightly towards relevance\n",
    "    top_k=10\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Original Top 10 Results:\n",
      "1. Blairism (score: 0.6392)\n",
      "2. Jonathan Powell (Labour adviser) (score: 0.6245)\n",
      "3. Vicky Pope (score: 0.6154)\n",
      "4. Tony Abbott (score: 0.6050)\n",
      "5. Earth in the Balance (score: 0.6040)\n",
      "6. Hadley Centre for Climate Prediction and Research (score: 0.5964)\n",
      "7. Jack Straw (score: 0.5962)\n",
      "8. Minister of Climate and the Environment (score: 0.5942)\n",
      "9. John Gummer (score: 0.5930)\n",
      "10. Abrupt climate change (score: 0.5891)\n",
      "\n",
      "MMR Reranked Top 10 Results:\n",
      "1. Blairism (score: 0.6392)\n",
      "2. Vicky Pope (score: 0.6154)\n",
      "3. Earth in the Balance (score: 0.6040)\n",
      "4. Jonathan Powell (Labour adviser) (score: 0.6245)\n",
      "5. Singaporean response to Hurricane Katrina (score: 0.5791)\n",
      "6. Abrupt climate change (score: 0.5891)\n",
      "7. Frank Sartor (score: 0.5891)\n",
      "8. Tony Abbott (score: 0.6050)\n",
      "9. United Nations Framework Convention on Climate Change (score: 0.5883)\n",
      "10. John Gummer (score: 0.5930)\n"
     ]
    }
   ],
   "source": [
    "# Compare original vs. reranked results\n",
    "print(\"\\nOriginal Top 10 Results:\")\n",
    "for i, doc in enumerate(results_list[:10], 1):\n",
    "    print(f\"{i}. {doc['title']} (score: {doc['@search.score']:.4f})\")\n",
    "\n",
    "print(\"\\nMMR Reranked Top 10 Results:\")\n",
    "for i, doc in enumerate(reranked_results, 1):\n",
    "    print(f\"{i}. {doc['title']} (score: {doc['@search.score']:.4f})\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note, for hybrid search and hybrid + semantic ranking, the same post-processing approach will apply."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate_mmr_diversity(results: List[Dict[str, Any]]) -> float:\n",
    "    \"\"\"Calculate average pairwise similarity as a diversity metric.\"\"\"\n",
    "    if len(results) <= 1:\n",
    "        return 1.0\n",
    "    \n",
    "    similarities = []\n",
    "    for i in range(len(results)):\n",
    "        for j in range(i + 1, len(results)):\n",
    "            sim = cosine_similarity(results[i]['embedding'], results[j]['embedding'])\n",
    "            similarities.append(sim)\n",
    "    \n",
    "    return sum(similarities) / len(similarities)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x600 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Numerical Comparison:\n",
      "------------------------------------------------------------\n",
      "Document                       Original Score  MMR Score      \n",
      "------------------------------------------------------------\n",
      "Blairism                                0.6392          0.6392\n",
      "Jonathan Powell (Labour advise          0.6245          0.6154\n",
      "Vicky Pope                              0.6154          0.6040\n",
      "Tony Abbott                             0.6050          0.6245\n",
      "Earth in the Balance                    0.6040          0.5791\n",
      "Hadley Centre for Climate Pred          0.5964          0.5891\n",
      "Jack Straw                              0.5962          0.5891\n",
      "Minister of Climate and the En          0.5942          0.6050\n",
      "John Gummer                             0.5930          0.5883\n",
      "Abrupt climate change                   0.5891          0.5930\n"
     ]
    }
   ],
   "source": [
    "import seaborn as sns \n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "def visualize_mmr_impact(original_results: list, mmr_results: list, figsize=(15, 6)):\n",
    "    \"\"\"\n",
    "    Create heatmap visualizations comparing original and MMR-reranked results using seaborn.\n",
    "    \n",
    "    Args:\n",
    "        original_results: List of original search results with embeddings\n",
    "        mmr_results: List of MMR reranked results with embeddings\n",
    "        figsize: Tuple for figure size\n",
    "    \"\"\"\n",
    "    # Calculate similarity matrices\n",
    "    def create_similarity_matrix(results):\n",
    "        n = len(results)\n",
    "        sim_matrix = np.zeros((n, n))\n",
    "        for i in range(n):\n",
    "            for j in range(n):\n",
    "                sim_matrix[i, j] = cosine_similarity(\n",
    "                    results[i]['embedding'],\n",
    "                    results[j]['embedding']\n",
    "                )\n",
    "        return sim_matrix\n",
    "\n",
    "    # Create the similarity matrices\n",
    "    orig_sim_matrix = create_similarity_matrix(original_results)\n",
    "    mmr_sim_matrix = create_similarity_matrix(mmr_results)\n",
    "    \n",
    "    # Create labels for the heatmaps\n",
    "    orig_labels = [doc['title'][:30] + '...' for doc in original_results]\n",
    "    mmr_labels = [doc['title'][:30] + '...' for doc in mmr_results]\n",
    "    \n",
    "    # Create the visualization\n",
    "    fig, (ax1, ax2) = plt.subplots(1, 2, figsize=figsize)\n",
    "    \n",
    "    # Plot original results heatmap\n",
    "    sns.heatmap(orig_sim_matrix, \n",
    "                ax=ax1,\n",
    "                cmap='Blues',\n",
    "                xticklabels=orig_labels,\n",
    "                yticklabels=orig_labels,\n",
    "                vmin=0,\n",
    "                vmax=1,\n",
    "                cbar_kws={'label': 'Cosine Similarity'})\n",
    "    ax1.set_title('Original Results Similarity')\n",
    "    plt.setp(ax1.get_xticklabels(), rotation=45, ha='right')\n",
    "    plt.setp(ax1.get_yticklabels(), rotation=0)\n",
    "    \n",
    "    # Plot MMR results heatmap\n",
    "    sns.heatmap(mmr_sim_matrix,\n",
    "                ax=ax2,\n",
    "                cmap='Blues',\n",
    "                xticklabels=mmr_labels,\n",
    "                yticklabels=mmr_labels,\n",
    "                vmin=0,\n",
    "                vmax=1,\n",
    "                cbar_kws={'label': 'Cosine Similarity'})\n",
    "    ax2.set_title('MMR Reranked Results Similarity')\n",
    "    plt.setp(ax2.get_xticklabels(), rotation=45, ha='right')\n",
    "    plt.setp(ax2.get_yticklabels(), rotation=0)\n",
    "    \n",
    "    # Calculate diversity scores (average off-diagonal similarity)\n",
    "    def calculate_diversity_score(matrix):\n",
    "        n = len(matrix)\n",
    "        off_diag_sum = (matrix.sum() - matrix.trace()) # Sum all except diagonal\n",
    "        off_diag_count = n * n - n # Number of off-diagonal elements\n",
    "        return off_diag_sum / off_diag_count if off_diag_count > 0 else 0\n",
    "    \n",
    "    orig_diversity = calculate_diversity_score(orig_sim_matrix)\n",
    "    mmr_diversity = calculate_diversity_score(mmr_sim_matrix)\n",
    "    \n",
    "    # Add text annotations with diversity scores\n",
    "    plt.figtext(0.3, 0.02, f'Diversity Score: {orig_diversity:.3f}', \n",
    "                ha='center', va='center')\n",
    "    plt.figtext(0.7, 0.02, f'Diversity Score: {mmr_diversity:.3f}', \n",
    "                ha='center', va='center')\n",
    "    \n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "    \n",
    "    # Print numerical comparison\n",
    "    print(\"\\nNumerical Comparison:\")\n",
    "    print(\"-\" * 60)\n",
    "    print(f\"{'Document':30} {'Original Score':15} {'MMR Score':15}\")\n",
    "    print(\"-\" * 60)\n",
    "    for orig, mmr in zip(original_results, mmr_results):\n",
    "        print(f\"{orig['title'][:30]:30} {orig['@search.score']:15.4f} {mmr['@search.score']:15.4f}\")\n",
    "\n",
    "# Use the visualization\n",
    "visualize_mmr_impact(results_list[:10], reranked_results)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Applying Maximum Marginal Relevance (MMR) in Retrieval-Augmented Generation (RAG)\n",
    "\n",
    "In Retrieval-Augmented Generation (RAG) scenarios, Maximum Marginal Relevance (MMR) balances relevance and diversity among retrieved documents. This approach is vital when the context window of a language model is limited, ensuring that every retrieved document provides unique and valuable information.\n",
    "\n",
    "**Key Benefits of MMR in RAG:**\n",
    "\n",
    "- **Mitigating Redundancy:**  \n",
    "  MMR reduces repetition by selecting documents that are both pertinent to the query and distinct from each other. This ensures the model’s limited context window isn't occupied by redundant details.\n",
    "\n",
    "- **Enhancing Comprehensiveness:**  \n",
    "  By incorporating a range of perspectives, MMR enables the creation of more thorough and nuanced responses, especially important in complex question-answering tasks.\n",
    "\n",
    "- **Optimizing Context Utilization:**  \n",
    "  When token constraints are tight, MMR ensures that each retrieved document contributes uniquely to the final answer, making full use of the available context space.\n",
    "\n",
    "**When to Consider MMR:**\n",
    "\n",
    "- **Limited Context Windows:**  \n",
    "  If your LLM has strict token limits, MMR helps ensure a diverse and relevant set of documents, maximizing the utility of the available context.\n",
    "\n",
    "- **Complex Information Needs:**  \n",
    "  For queries that demand multifaceted answers, MMR surfaces documents covering various angles, enabling more informed and holistic responses.\n",
    "\n",
    "- **Avoiding Redundancy:**  \n",
    "  When initial retrieval results contain similar or duplicate content, MMR promotes diversity, boosting the overall quality and depth of generated outputs.\n",
    "\n",
    "In summary, MMR is a powerful tool in RAG pipelines. It carefully curates a diverse set of relevant documents, leading to richer, more comprehensive, and ultimately more useful generated responses.\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}