Term-Based Retrieval

Suppose you type a few words into a search box and get back a list of results. Somewhere behind that list is a decision the system had to make for every item it could show you: how well does this item match what I asked for? Retrieval is the name for the family of techniques that make that decision and act on it. This article introduces the idea, shows where it sits in the larger picture, and sets up the two mechanisms that the rest of this chapter builds on.

At its core, retrieval works by ranking documents based on their relevance to a given query. The important word is ranking. Every retrieval method has to answer the same underlying question, which is how relevant each document is to the query, and then sort on the answer. The result is an ordered list, not a set of accept and reject decisions.

The operation has a clear shape. The query and the document collection both feed into a step that computes a relevance score for each document, and those scores determine the final order. The step that changes from one retrieval method to another is the middle one, where the score is computed. Everything before it and after it stays the same.

It is easy to associate retrieval with retrieval-augmented generation (RAG), because RAG pipelines depend on fetching relevant material before a language model writes its answer. But retrieval is not unique to RAG. Information retrieval is an old idea, and it already forms the backbone of search engines, recommender systems, and log analytics. These systems existed long before RAG did.

The practical consequence is worth remembering. Many retrieval algorithms were developed for traditional retrieval systems, and those same algorithms can also be reused inside RAG. When you meet a retrieval method in a RAG setting, it may have a long history in search and recommendation work behind it.

Two words appear constantly in this area, and they are not quite the same. Retrieval is typically limited to one database or system. Search involves retrieval across various systems. The difference is one of scope: retrieval describes finding relevant items within a single source, while search describes the wider act of looking across several sources.

The chapter introduces two common retrieval mechanisms as a starting point: term-based retrieval and embedding-based retrieval. Each one is a different answer to the scoring question raised earlier. Both answer the same question of how relevant each document is to the query, and they differ in how they compute the score that drives the ranking.

The central lesson is that the goal stays fixed while the scoring method varies. Term-based and embedding-based retrieval are two answers to one question. A broader way of grouping such methods, sparse versus dense retrieval, comes in the next part of the chapter, building on these scoring differences.

A term-based method has to connect each query term to the documents that contain it. The sketch below is an illustrative picture of one way to do that. It is not a detail the chapter specifies. Each term on the left is linked to the document IDs where it appears, so a query term can be followed directly to its documents rather than checking every document one by one.

Read the sketch as a toy illustration. If the query is rain umbrella, the query terms point to doc 2 and doc 5, so those are the documents a term-based method would have to score. The term lookup narrows the work, and the scoring step then decides how relevant each candidate is.

The two starting mechanisms can be lined up side by side. Both aim at the same outcome, a ranked list ordered by relevance to the query. The table below separates them by the feature they differ on, which is how each one computes its relevance score.