Browse Prior Art Database

Method and System for Presenting Structured Ads in Sponsored Search

IP.com Disclosure Number: IPCOM000200153D
Publication Date: 2010-Oct-01
Document File: 4 page(s) / 31K

Publishing Venue

The IP.com Prior Art Database

Related People

Anand Murugappan: INVENTOR [+2]

Abstract

A method and system is provided for presenting structured ads in a sponsored search. Query template candidates are automatically generated for placing a structured ad corresponding to a search query. A Structured Ad Module (SAM) for placing structured ads is created and metadata relevant to the structured ad is provided.

This text was extracted from a Microsoft Word document.
This is the abbreviated version, containing approximately 45% of the total text.

Method and System for Presenting Structured Ads in Sponsored Search

Abstract

A method and system is provided for presenting structured ads in a sponsored search.  Query template candidates are automatically generated for placing a structured ad corresponding to a search query.  A Structured Ad Module (SAM) for placing structured ads is created and metadata relevant to the structured ad is provided.

Description

Disclosed is a method and system for presenting structured ads in a sponsored search.

In general, structured information is easier to be processed and acted upon by users.  Presenting users with ads in a structured format improves user experience and enables search engine providers to efficiently use valuable real estate on a search result webpage.

The method and system disclosed herein, automatically generates query template candidates for placing a structured ad corresponding to a search query.  For example, the most frequent query templates may be identified based on search logs such as, %FOOD ITEM% recipes, %LOCATION% things to do, %MAKE% %MODEL% %YEAR% or %AIRPORT% to %AIRPORT% etc.  Given the search query, a segmentation system such as for example, a CRF text segmentation system identifies segments and assigns labels based on ontology.  For example, the segmentation system identifies "Bank of America" as a segment in "Bank of America Jobs" and "Bank of America" as %ORGANISATION NAME%.  However, if the query were "banks in America", the segmentation system identifies the segment "banks" as %ORGANISATION% and the segment "America" as %COUNTRY NAME%.  Further, individual queries in query logs are segmented to produce query templates.  For example, a search query "Used 2007 Toyota* Camry*" may be segmented as "Used", "2007", "Toyota", "Camry" and then cast into a template to form "Used, %YEAR% %MAKE% %MODEL%".

Thereafter, the generated query templates are examined and the fields that are to be requested from the advertisers are determined.  The method further involves, enforcing data types corresponding to the fields and defining rules to semi-automatically validate data provided by advertisers.  For example, the value for "year" needs to be an integer type.  Subsequently, at serve time, a Structured Ad Module (SAM) is created in a predefined format for placing structured ads.  The SAM may be created in different shapes and sizes based on a real-estate available on a search result webpage.  Query templates, such as %WEBSITES% may not be suitable for a SAM module.  For example, the query template %WEBSITES% is unlikely to have bidders or offer comparable data since these might be registered trademarks.

Subsequently, advertisers are recommended a query template in response to a bidded term.  When an advertiser enters the bidded term, the system segments the bidded term and recommends available query templates to the advertiser.  The advertiser may then commit to a suggested query template and provide metadata relevant t...