Teragram

OEM Products

 

Related Queries

Teragram Related Queries is a feature offering related search suggestions that is fast becoming a recognized standard in today's search applications. For example, if someone is searching for "shoes," the system might suggest the following related searches, "Running Shoes, Leather Shoes, Suede Shoes, Tennis Shoes, Nike Shoes," among others. This feature can be used to suggest more specific related searches for your text or Ecommerce search engines. Catalog and Ecommerce search engines use this feature to increase their traffic, improve sales, and to make customers aware of promotions of a wider range of products.

Teragram Telated Queries uses the implicit intelligence of a large querying audience; it produces related queries that will guide and enrich the search experience for all users. Related Search is a software package that enables any system to make the necessary statistical correlations between past queries from large unsorted query logs. This enables the precise generation of related searches for any future query. This is a valuable data-mining process that is fast, easy-to-use, and critically controllable. Related searches can be instantly returned from within any application using Teragram's robust internal interface. In addition to statistical log analysis, Teragram's Related Queries software uses hand-crafted data that allows related pre-determined searches for sponsored links or for other special promotions.

Usage of Related Search from Teragram is comprised of two distinct phases. First, a compiler supplied by Teragram will statistically process a raw query log to produce a compressed data structure known as a Teragram classifier. Teragram classifiers are efficient and fast, and this one will be customized with the input data set for the production of related searches. In an easy second step, the classifier data is then made available to a calling application through the Teragram Classifier API (Application Programming Interface). Now any search application can instantly generate the related searches for new query inputs.

Back to Top