“Did you mean” (DYM) is a search feature that suggests a corrected query when the input appears to be misspelt or unusual. It relies on similarity measures between words, language models and analysis of frequent searches. Example: someone typing “coffe machne” into an online shop sees “Did you mean: coffee machine?”. The feature reduces empty result pages and thus improves the search experience.
Fuzzy matching finds entries that are similar but not exactly identical – for example because of typos, abbreviations or different spellings. It relies on similarity measures such as the Levenshtein distance, which counts how many characters need to be changed. A typical use is finding duplicates in address and customer data: ‘Smith Ltd, 5 High St’ and ‘Smith Limited, 5 High Street’ are recognised as probably the same record.
Query expansion automatically adds related terms to a search query so that relevant documents using different wording are also found. Typical additions are synonyms, abbreviations, word forms or translations: a search for “cancellation” might then also find texts about “termination of contract”. This is especially useful for keyword search, but it can reduce precision if too many or unsuitable terms are added. Today, language models are often used to generate the expanded queries.
AI search refers to search methods that use AI to grasp the intent behind a query, weight results and often generate a summarised answer with sources rather than a plain list of links. Technically, they use keyword search, vector search over embeddings or a combination, frequently followed by a language model. Examples include answer engines on the web and internal knowledge search that helps staff find what they need in manuals and policies.
Information retrieval covers methods that find the documents or passages in a large body of text that match a query and rank them by relevance. Classic approaches such as BM25 weight search terms by how often and how rarely they occur; newer ones add semantic similarity. It underpins search engines, enterprise search and retrieval-augmented generation, in which a language model uses the retrieved passages as the basis for its answer.
