Monday, January 06, 2014

Understanding title information shown in SharePoint 2013 search results and how to make it work better

The title field which is shown in the SharePoint 2013 search results is by default mapped to the crawled properties as outlined in the table. There is however a hidden mechanism present in SharePoint 2013 which is called  title metadata extraction which overwrite the “Title” managed property with a value SharePoint extracts from parts of Word and PowerPoint documents (See SharePoint 2013 Ceres Shell – down the search rabbit hole for more details). In SharePoint 2010 you could override this behavior by changing the EnableOptimisticTitleOverride registry key but unfortunately this does not work in SharePoint Server 2013.

Crawled Property Origin Description
TermTitle SharePoint According to the documentation this is the title of an item in SharePoint but in my environment it is not filled in for any item. In my experience you can move this crawled property down in the priority for the managed property
Office:2 Office Office:2(Text) crawled property maps onto Title property which is used in the Word document itself as part of the summary information
ows_BaseName SharePoint Name of the SharePoint page
Title DocParser Title as picked up by the content processing component
MailSubject DocParser The subject of an email file as picked up by the content processing component
Mail:5 Mail Subject line of an email file
People:PreferredName People PreferredName property in the SharePoint user profile
Basic:displaytitle Basic Filename of Office document
ows_Title SharePoint Maps to SharePoint list field Title
Basic:10 Basic Mapped to Title,FileName
Basic:9 Basic Path metadata
If you install the SharePoint Server 2013 October 2013 Cumulative Update, you will notice that a new property is added called MetadataExtractorTitle. This property is filled in by the Microsoft.MetadataExtractorSubFlow.dll processing flow.



So if you want to be sure that the title of the an item is used in SharePoint search result, you can move the ows_Title crawled property before the MetadataExtractorTitle property. If one of the crawled properties is empty, the next crawled property will be used. This means that if the user does not specify a title in SharePoint, SharePoint will try to use metadata extraction for Word and PowerPoint files to give a an alternative for the title.
Unfortunately, you will need to perform an index reset first and afterwards you will need to do a full crawl to see the result of your changes.
References:

No comments:

Post a Comment