OSINT

Eyetracking studies show that users sometimes look at only a single result on a search-results page because that result is good enough for their needs.

A sandbox of search design illustrations, including a pattern library for search and discovery.

How to learn similar terms in a given unsupervised corpus using Word2Vec

How data splitting can be done and why it is important for the offline evaluation of Learning to Rank models?

People want to buy things that have many great reviews. But when sorting by average rating, the best results are buried by things that have one or two perfect reviews. You can solve this problem by ba...

Transparent and trustworthy way to support your good cause by simply searching the Internet.

There is one topic that gives even the toughest search engineers a headache, multi-word synonyms. There are some ways to sort of get a solution, one of them is a nice tool called Querqy. Querqy is created by René Kriegler....

Efficient topic modelling in Python

Because today’s search-results pages have many possible complex layouts, users don’t always process search results sequentially. They distribute their attention more variably across the page than in the past.

The most immediate context for a search query is a search session, a sequence of activities the searcher performs in order to pursue an…

Text Mining — Extracting Synonyms and Antonyms

Spelling correction is a must-have for any modern search engine. Estimates of the fraction of misspelled search queries vary, but a variety…

Different forms of a word often communicate essentially the same meaning. For example, there’s probably no difference in intent between a…

How to use synonyms and synonym filters in Elasticsearch. Synonyms are a powerful tool for increasing the recall of your search system, but there are many subtleties that are important to know and exp...

An effective way of reducing the dimensionality of your data

Whether it’s the time of year or the time of day, the time when someone performs a search sometimes help us determine the searcher’s…

Search result snippets, also known as query-biased summaries, are the additional context included with each result on the search results…

In order to understand queries, it’s important to ground that understanding in a knowledge base. Two common ways to represent a knowledge…

Search queries that express broad intent often return intractably large result sets. We’ve already discussed faceted search as a way to…

Until now, we’ve mostly focused on query processing — which is to be expected, given that this series is about query understanding. But…

Information Retrieval (IR) systems are a vital component in the core of successful modern web platforms. The main goal of IR systems is to provide a communication layer that enables customers to establish a retrieval dialogue with underlying data.

Introduction

Conventional wisdom and textbooks say BK-trees are especially suited for spelling correction and fuzzy string search. But does this really…

Most search applications assume a query-response paradigm: the searcher submits a search query, and the search engine responds with…

Now that we can handle characters, let’s move on to words.

Faceted search is a topic broad enough to deserve its own book. It has become a standard feature of all modern search engines, including…

Location often provides a strong signal of searcher intent. Sometimes location acts as an implicit part of the search query, such as when…

In English Language, people generally type the queries which are separated by space, but sometimes and somehow this space is found to be…