Search Overview + Documentation

Follow

Overview

Synthesis Search uses query federation to retrieve search results from both the Synthesis content database and the SharePoint site collection. User queries are passed simultaneously to the Synthesis Search Engine (an implementation of Apache Lucene.Net) and to SharePoint Search (using the SharePoint 2013 Query API), and the returned results are combined in the client and presented via the Synthesis Search Results page.

Refer to the Data Dictionary for a list of Synthesis employee, project, content, company, and opportunity database fields and their relationship to Deltek Vision source fields.

The following content categories are searched by the Synthesis Search Engine:

 And the following content categories are searched in Microsoft SharePoint: 

Refer to Appendix A for more details regarding Knowledge Architecture’s Microsoft SharePoint 2013 Search integration and implementation.

Security/Permissions

We rely on the SharePoint API to enforce permissions for queries against content stored in the SharePoint site collection. For content stored in the Synthesis database, the permission levels configured in Synthesis are similarly respected. If an Employee does not have permission to see a Post or a Community or an Opportunity, those items will not be returned by the Synthesis Search Engine.

Search Operators

Synthesis Search supports standard search operators such as AND, OR, NOT, Wildcards*, and "Exact Phrases". Synthesis Searches with more than one term use AND by default. More information on this topic can be found in the Search Operators article.

Employees

What Data Gets Indexed and Searched for Employees?

Only Employees in the Employee Directory are indexed by the Synthesis Search Engine.

The following fields are indexed but not searched:

  • EmployeeID, PrimaryImageID

The following string fields are indexed and searchable:

  • BillingCategory, City, Country, Department, Discipline, Email, EmployeeAdvocate, EmployeeNumber, EmployeeType, FirmName, FirstName, FullName, LastName, MobilePhone, Office, OrgName, PersonalBio, PreferredName, ProjectRoles, ProjectTypes, ResourceFor, State, Status, StreetAddress1, StreetAddress2, StreetAddress3, StreetAddress4, Title, WorkPhone, WorkPhoneExtension, WorkPhonePlusExtension, ZIP, ProfessionalBio 

The following Yes/No type fields are indexed and searchable (Refer to Appendix B for methodology):

  • AIA, LEEDAccredited, NCARB, ProfessionalEngineer, RegisteredArchitect

In addition, all mapped varchar(255) or varchar(max) custom fields that have been given display names are both indexed and searchable. Mapped custom fields containing Yes/No content are indexed and searchable as indicated in Appendix B.

The following Synthesis grid fields are indexed and searchable if mapped:

  • KA_EmployeeSkills grid - Skill
  • KA_EmployeeCertifications grid - Certification, CertifyingAgency
  • KA_EmployeeResumes grid - ResumeText
  • KA_EmployeeDegrees grid - Degree, Specialty, Institution
  • KA_EmployeeProfessionalAffiliations grid - Affiliation
  • KA_EmployeeRegistrations grid - Registration

How are Search Results Ranked For Employees?

Search Terms matching the Employee’s FirstName, LastName, PreferredName, or FullName fields will be boosted by a factor of 1. Search Terms matching the EmployeeTitle field or the Skill field in the Employee Skills Grid will be boosted by a factor of .5. All other fields will be scored identically. 

Search Results for Employees will be sorted by Search Rank Score descending, then alphabetically by LastName, and then FirstName.

What Search Results Facets are Available for Employees?

Office and Title

 

Projects

What Data Gets Searched for Projects?

Only Projects in the Project Directory are indexed by the Synthesis Search Engine. In addition, grid data must also be marked “Show in Directory” in order to be indexed.

The following fields are indexed but not searched:

  • ProjectID, PrimaryImageID, ClientID, ClientInDirectory

The following string fields are indexed and searchable:

  • ProjectNumber*, City, ClientName, Contractor, Country, LEEDCertificationLevel, LEEDCertificationStatus, LEEDProgram, MarketingName, MarketingTier, Office, OrgName, ParentProjectNumber*, PIC, ProjectAliases, ProjectDeliveryMethod, ProjectManager, ProjectName, ProjectPhase, ProjectRole, ProjectType, State, Status, StreetAddress1, StreetAddress2, StreetAddress3, StreetAddress4, ZIP, OpportunityName

*These fields will be indexed both with existing punctuation and without, to enable “prefix-searching” - i.e. searching on 01045 would return Project Number 01045.005

The following Yes/No type fields are indexed and searchable (Refer to Appendix B for methodology):

  • PublicSector, LEEDProject, ParentProject

In addition, all mapped varchar(255) or varchar(max) custom fields in the KA_Projects table that have been given display names are both indexed and searchable. Mapped custom fields containing Yes/No content are indexed and searchable as indicated in Appendix B.

The following project grid fields are indexed and searchable: 

  • KA_ProjectAwards grid - Description, Institution
  • KA_ProjectDescriptions grid - Description
  • Project keyword values imported from OpenAsset, if applicable 

How are Search Results Ranked for Projects?

Search Terms matching ProjectName, MarketingName, ProjectNumber, or ProjectAliases fields will be boosted by a factor of 1. Search Terms matching the ClientName field will be boosted by a factor of .5. All other fields will be scored identically.

Search Results for Projects will be sorted by Search Rank Score descending and then alphabetically by ProjectName.

What Search Results Facets are Available for Projects?

Client Name, Status, and Project Type

 

Contacts

What Data Gets Searched for Contacts?

Only Contacts in the Contact Directory are indexed by the Synthesis Search Engine.

The following fields are indexed but not searched:

  • ContactID, PrimaryImageID, CompanyID, CompanyInDirectory

The following string fields are indexed and searchable:

  • CompanyName, CompanyPrimaryDiscipline, ContactFirstName, ContactLastName, ContactName, ContactPreferredName, ContactType, Department, Email, MobilePhone, ProjectRoles, Status, CompanyOfficePhone, Title, WorkPhone

The following fields are indexed and searchable for each of a company’s locations:

  • AddressDesc, City, Country, Fax, Phone, State, StreetAddress1, StreetAddress2, StreetAddress3, StreetAddress4, ZIP

In addition, all mapped varchar(255) or varchar(max) custom fields that have been given display names are both indexed and searchable. Mapped custom fields containing Yes/No content are indexed and searchable as indicated in Appendix B. 

How are Search Results Ranked for Contacts?

Search Terms matching ContactFirstName, LastName, PreferredName, or FullName fields will be boosted by a factor of 1. Search Terms matching CompanyName field will be boosted by a factor of .5. All other fields will be scored identically. 

Search Results for Contacts will be sorted by Search Rank Score descending, then alphabetically by LastName, and then FirstName.

What Search Results Facets are Available for Contacts?

Company, Title, and City

 

Companies

What Data Gets Searched for Companies?

Only Companies in the Company Directory are indexed by the Synthesis Search Engine.

The following fields are indexed but not searched:

  • CompanyID, PrimaryImageID

The following string fields are indexed and searchable:

  • CompanyName, CompanyNumber, CompanyType, DisciplineDescription, Email, Notes, PrimaryDiscipline, ProjectRoles, RelationshipStatus, Status, Website

The following Yes/No type fields are indexed and searchable (Refer to Appendix B for methodology):

  • DisabledVetOwnedSmallBusiness, DisadvBusiness, HBCU, MinorityBusiness, PriorWork, Recommend, SmallBusiness, VetOwnedSmallBusiness, WomanOwned

The following fields are indexed and searchable for each of a company’s locations:

  • AddressDesc, City, Country, Fax, Phone, State, StreetAddress1, StreetAddress2, StreetAddress3, StreetAddress4, ZIP

In addition, all mapped varchar(255) or varchar(max) custom fields that have been given display names are both indexed and searchable. Mapped custom fields containing Yes/No content are indexed and searchable as indicated in Appendix B.

How are Search Results Ranked for Companies?

Search Terms matching CompanyName field will be boosted by a factor of 1. All other fields will be scored identically.

Search Results for Companies will be sorted by Search Rank Score descending then alphabetically by Company.

What Search Results Facets are Available for Companies?

Company Type and Primary Discipline

 

Opportunities

What Data Gets Searched for Opportunities?

Only Opportunities in the Opportunity Directory are indexed by the Synthesis Search Engine. 

The following fields are indexed but not searched:

  • OpportunityID, PrimaryImageID, ClientID, ClientInDirectory

The following string fields are indexed and searchable:

  • AssocProjectName, AssocProjectNumber*, AssocPromotionalProjectName, AssocPromotionalProjectNumber*, City, ClientName, ContactName, Country, Description, OpportunityName, OpportunityNumber*, OpportunityType, OrgName, PIC, ProjectManager, Source, Stage, State, Status, StreetAddress1, StreetAddress2, StreetAddress3, StreetAddress4, Zip, Office, OpportunityRole

*These fields will be indexed both with existing punctuation and without, to enable “prefix-searching” - i.e. searching on 01045 would return Opportunity Number 01045.005

In addition, all mapped varchar(255) or varchar(max) custom fields that have been given display names are both indexed and searchable. Mapped custom fields containing Yes/No content are indexed and searchable as indicated in Appendix B. 

How are Search Results Ranked for Opportunities?

Search Terms matching OpportunityName and OpportunityNumber fields will be boosted by a factor of 4. Search Terms matching ClientName field will be boosted by a factor of 2. All other fields will be scored identically.

Search Results for Opportunities will be sorted by Search Rank Score descending and then alphabetically by OpportunityName.

What Search Results Facets are Available for Opportunities?

Client Name, State, and Opportunity Type

 

Communities

What Data Gets Searched for Communities?

The following fields are indexed but not searched:

  • CommunityID, PublicSite

The following string fields are indexed and searchable:

  • CommunityName, CommunityUrl

How are Search Results Ranked for Communities?

Search Terms matching CommunityName field will be boosted by a factor of 1.

Search Results for Communities will be sorted by Search Rank Score descending and then alphabetically by CommunityName.

What Search Results Facets are Available for Communities?

None

 

Topics

What Data Gets Searched for Topics?

All Topics (also known as hashtags) will be searchable unless disabled in the Hashtag Administration web part on the Stream Administration page.

The following fields are indexed and searchable:

  • Hashtag

NOTE: Searching with and without the preceding “#” will return matching Topics 

How are Search Results Ranked for Topics?

Search Results for Topics will be sorted by Search Rank Score descending and then alphabetically by Hashtag. 

What Search Results Facets are Available for Topics?

None

 

Posts

Posts consist of items in the Synthesis activity stream. 

What Data Gets Searched for Posts?

The following fields are indexed but not searchable for Posts:

  • ItemID, PublicSite, DateStamp (post date)

The following fields are indexed and searchable for Posts:

  • AuthorName (post author name), Title, FormattedContent (post body), SubItemBody (attached item body), SubItemTitle (attached item title)

In addition, the following data associated with Posts is indexed and searchable:

  • Associated Comments - AuthorName (comment author name), FormattedContent (comment body)
  • Associated Tags - EntityName (@mentions)
  • Associated Topics (i.e. Hashtags) - Hashtag 

How are Search Results Ranked for Posts?

Search Terms matching Title field will be boosted by a factor of 4.

Search Terms matching FormattedContent (post and comments), SubItemTitle, and SubItemBody will be boosted by a factor of 2.

Post Freshness will also adjust the Search Rank Score by a factor of 2. (See below.)

Search Terms matching AuthorName (post and comments), EntityName, and Hashtag fields will be boosted by a factor of 1. 

What is Freshness?

Newer posts are ranked (not just sorted) higher than older posts.

How are Search Results Sorted for Posts?

Search Results for Posts will be sorted by Search Rank Score and then reverse chronologically by DateStamp.

What Search Results Facets are Available for Posts?

My Activity, Posted Date, Author, Community, and Topic (Hashtag)

 

Documents

For the purposes of Synthesis 5 Search, a Document is a file in a Microsoft SharePoint list or library. The file does not have to be stored in a “Document Library” in order to be considered a Document. Files will be returned under the heading of Documents as long as they are not aspx pages or html documents (those results will appear in “Other Results”), and they do not have file extensions that identify them as Images or Videos.

What Data Gets Searched for Documents?

We use the Microsoft SharePoint Query API to search:

  • Document Content
  • Document Metadata including all text values from SharePoint List Columns, including Filename, Title, and Custom Columns
  • For a complete list of metadata crawled by SharePoint 2013, please see the following documentation from Microsoft: https://technet.microsoft.com/en-us/library/JJ219630.aspx

In addition, when a user query is submitted, Synthesis 5 Search traverses SharePoint navigation links, which are typically not crawled by SharePoint’s search engine, looking for any matching links that point to internal documents. In the event a match is found, that document is returned - for example, if a navigation item in the Quick Links of a site called “Benefits Provider” links to a document that does not contain the word “benefits” in any crawled property, that document would still be returned by a search for the word “benefits”.

Note: SharePoint folders do not act as metadata for the documents within them. 

How are Search Results Ranked for Documents?

Search Results for Documents will be sorted by the Search Rank Score returned by the SharePoint API, descending. For more information on the SharePoint 2013 Search Ranking Model, please see Appendix D.

What Search Results Facets are Available for Documents?

Document Type, Uploaded Date, and Community

 

Images

Image results are federated from both Microsoft SharePoint and Synthesis Search. Microsoft SharePoint Search returns images that are contained in Microsoft SharePoint lists or libraries - in order to be returned, they must have one of the file extensions listed in Appendix C and must be over 5kb in size (in order to avoid returning results of negligible value, such as icons). Synthesis Search returns images that have been imported from Axomic OpenAsset or that have been uploaded by users to the activity stream. 

What Data Gets Searched for Images?

For Microsoft SharePoint Images:

We use the Microsoft SharePoint Query API to search images contained in Microsoft SharePoint. For a detailed description of the Keyword Query text passed to the API, refer to Appendix C. In addition, in the event a search query matches the title of a wiki page, we will return in the image results any images embedded in the content of that page.

For Synthesis Images:

The following fields are indexed but not searchable:

  • MediaID, OriginalHeight, OriginalWidth, DateUploaded, EmployeeID (if an employee image), ProjectID (if a project image), ItemID, PublicSite

The following fields are indexed and searchable:

  • Caption (OpenAsset image only), Description (OpenAsset image only), OriginalFilename; FullName of Employee (if an employee image); ProjectNumber, MarketingName (of associated project), and ClientName (of associated project)
  • ProjectNumber, MarketingName, ClientName, EmployeeID, FullName -- If the image is associated with a Project or Employee

The following associated information is also indexed and searchable for images coming from Axomic OpenAsset:

  • ImageKeywords
  • ProjectKeywords (If the image is associated with the project in Axomic OpenAsset)
  • AlbumNames, AlbumDescriptions, AlbumKeywords (If the image belongs to an album or albums)

The following associated information is also indexed and searchable for images uploaded to Synthesis with Share Something:

  • AuthorName, Title, FormattedContent, SubItemTitle, SubItemSubtitle, SubItemBody

How are Search Results Ranked for Images?

For Microsoft SharePoint Images: By Microsoft SharePoint Search. 

For Synthesis Images: Search Terms matching OriginalFilename, Caption, ProjectNumber, MarketingName (of associated project), and FullName of Employee (if an employee image) will be boosted by a factor of 3. Search Terms matching the Description, ProjectKeywords, ImageKeywords, and AlbumNames, will be boosted by a factor of 2. Search Terms matching AlbumDescriptions, AlbumKeywords and ClientName will be boosted by a factor of 1. All other fields will be scored identically.

For Images Associated with Microsoft SharePoint Blog Posts or Synthesis Posts: Search Terms matching Title, AuthorName (post and comment authors), SubItemTitle, EntityName (@Mentions), and Hashtags will be boosted by a factor of 2. Search Terms matching FormattedContent, SubItemTitle, SubItemBody, and FormattedContent (comment body) will be boosted by a factor of 1. All other fields will be scored identically.

Search Results for SharePoint Images, Synthesis Images, and Images Associated with Microsoft SharePoint Blog Posts or Synthesis Posts are sorted by their Search Rank Score, then Search Rank Scores from all three categories of images are normalized and combined into a single result set.

What Search Results Facets are Available for Images?

Image Source, Image Size, and Uploaded Date

 

Videos

What Content Sources are Searched for Videos?

  • Axomic OpenAsset Videos which meet Synthesis Import Criteria*
  • Microsoft SharePoint Asset Libraries
  • Other Microsoft SharePoint Libraries (No Video Preview Available)
  • Videos embedded in Synthesis Posts

NOTE: Axomic OpenAsset videos’ thumbnail image, URL, and associated metadata is imported from Axomic OpenAsset to Synthesis. Axomic OpenAsset videos files themselves are not imported from Axomic OpenAsset into Synthesis. 

What Data Gets Searched for Videos?

For Microsoft SharePoint Videos: We use the Microsoft SharePoint Query API to search videos contained in Microsoft SharePoint. 

For Synthesis Videos:

The following fields are indexed but not searchable:

  • MediaID, OriginalHeight, OriginalWidth, DateUploaded, EmployeeID (if an employee image), ProjectID (if a project image), ItemID, PublicSite

The following fields are indexed and searchable:

  • Caption (OpenAsset image only), Description (OpenAsset image only), OriginalFilename; FullName of Employee (if an employee image); ProjectNumber, MarketingName (of associated project), and ClientName (of associated project)
  • ProjectNumber, MarketingName, ClientName, EmployeeID, FullName -- If the video is associated with a project or employee

The following associated information is also indexed and searchable for videos in Axomic OpenAsset:

  • Image Keywords
  • Project Keywords -- If the video is associated with the project in Axomic OpenAsset
  • Album Names, AlbumDescriptions -- If the video belongs to an album or albums

The following associated information is also indexed and searchable for videos shared on Synthesis via a Share Something attachment:

  • AuthorName, Title, FormattedContent, SubItemTitle, SubItemSubtitle, SubItemBody

How are Search Results Ranked for Videos?

For Axomic OpenAsset Videos: Search Terms matching Original Filename and Video Caption will be boosted by a factor of 3. Search Terms matching Video Description, Keywords, and Album Names will be boosted by a factor of 2. Search Terms matching Album Descriptions and Album Keywords will be boosted by a factor of 1. All other fields will be scored identically.

For Videos Embedded in Synthesis Posts:  Search Terms matching Post Title, Post Author Name, Attached Item Title, Comment Author Name, @Mentions, and Topics (Hashtags) will be boosted by a factor of 2. Search Terms matching Post Body, Attached Item Subtitle, Attached Item Body, and Comment Body will be boosted by a factor of 1. All other fields will be scored identically.

For Videos Embedded in Synthesis Posts:  Search Terms matching Attached Item Title and Attached Item Subtitle boosted by a factor of 3. Search Terms matching Post Title, Post Author Name, @Mentions, and Topics (Hashtags) will be boosted by a factor of 2. Search Terms matching Post Body and Comment Body will be boosted by a factor of 1. All other fields will be scored identically.

Search Results for Videos will be ranked by Search Rank Score descending.

What Search Results Facets are Available for Videos?

Video Source and Uploaded Date

 

Wiki Pages

What Content Sources are Searched for Wiki Pages?

Microsoft SharePoint Wiki Page Libraries

What Data Gets Searched for Wiki Pages?

We use the Microsoft SharePoint Query API to search:

  • Wiki Page Content
  • Wiki Page Metadata including all text values from SharePoint List Columns, including Title and Custom Columns
  • For a complete list of metadata crawled by SharePoint 2013, please see the following documentation from Microsoft: https://technet.microsoft.com/en-us/library/JJ219630.aspx

In addition, when a user query is submitted, Synthesis 5 Search traverses Microsoft SharePoint navigation links, which are typically not crawled by Microsoft SharePoint’s search engine, looking for any matching links that point to wiki pages. In the event a match is found, that wiki page is returned

How are Search Results Ranked for Wiki Pages?

Search Results for Wiki Pages will be ranked by Search Rank Score descending. For more information on the SharePoint 2013 Search Ranking Model, please see Appendix D.

What Search Results Facets are Available for Wiki Pages?

Last Modified and Community

 

Events

What Content Sources are Searched for Events?

Microsoft SharePoint Calendars

What Data Gets Searched for Events?

We use the Microsoft SharePoint Query API to search calendars. 

How are Search Results Ranked for Events?

Search Results for Events will be ranked by Search Rank Score and then reverse chronologically by Event Date.

What Search Results Facets are Available for Events?

Time Frame and Community

 

Links

What Content Sources are Searched for Links?

Microsoft SharePoint Link Libraries, Top Navigation, and Left Navigation

What Data Gets Searched for Links?

We use the Microsoft SharePoint Query API to search link libraries. In order to be returned as a link result, the link must be to an external resource. Links to Microsoft SharePoint resources, such as Documents, Images and Wiki Pages, will be returned under those headings.

In addition, Synthesis Search returns results from Microsoft SharePoint navigation links. Both top and left (i.e. quick link) navigation items are searched. As with link libraries, in order to be returned under the “Links” header, it must be a link to an external resource.

How are Search Results Ranked for Links?

Search Results for Links will be ranked by Search Rank Score descending.

What Search Results Facets are Available for Links?

Community

 

Other Results 

These Microsoft SharePoint results include aspx or html pages, generic list items, and containers for items such as folders or lists.

What Data Gets Searched for Other Results?

We use the Microsoft SharePoint Query API to search all other content types not listed above. 

In order to avoid returning Microsoft SharePoint lists every time one of the items contained in that list is matched, the title of a list or library must match the query in order to be returned.

How are Search Results Ranked for Other Results?

Search Results for Other Results will be sorted Search Rank Score descending. For more information on the SharePoint 2013 Search Ranking Model, please see Appendix D.

What Search Results Facets are Available for Other Results?

Community

 

APPENDIX A: Microsoft SharePoint 2013 Search Integration

As mentioned in the overview, Knowledge Architecture uses query federation to return results from both the Synthesis Content Database and the Microsoft SharePoint Site Collection. We rely on Microsoft SharePoint 2013’s built-in search capabilities to crawl, index and query the content located in the site collection. Synthesis Search integrates with Microsoft SharePoint Search using the Microsoft SharePoint 2013 Query API.

For more information on the Microsoft SharePoint 2013 Query API: http://msdn.microsoft.com/EN-US/library/office/dn423226(v=office.15).aspx

Knowledge Architecture builds its SharePoint search queries using the Keyword Query Language (KQL) syntax:

http://msdn.microsoft.com/EN-US/library/office/ee558911(v=office.15).aspx

As with an out-of-the-box Microsoft SharePoint search implementation, all content sources, crawl schedules, and query rules are configured within Microsoft SharePoint Central Administration. Note that, while all content sources configured in Central Administration will be crawled and indexed, Knowledge Architecture’s query syntax is specifically designed to exclude certain results that may be duplicative of Synthesis Content Database results or that may provide a suboptimal search experience. For example, Microsoft SharePoint Blog Posts, Comments and Categories are excluded from the Microsoft SharePoint result set, since such results will be returned by the Synthesis Search Engine as part of its “Posts” category. Likewise, since Microsoft SharePoint Sites are synonymous with Synthesis Communities (which are returned via the Synthesis Search Engine), Microsoft SharePoint Sites are removed from the Microsoft SharePoint result set. Another example is that in order for Microsoft SharePoint Lists and Libraries to be returned as results, we require that the title of the List or Library (or at least the title of a List or Library view) contain the search term - without this restriction, we have found that List and Library views tend to overwhelm the result set.

Please note that any Search Scopes configured in Microsoft SharePoint Central Administration will be ignored by Synthesis Search. In addition, content sources other than Local Microsoft SharePoint Sites, such as external sites or BCS entities, may not be returned by our standard queries or, if returned, results from those sources may not be formatted properly. If you wish to add custom content sources to your Synthesis Search implementation, we recommend contacting KA Support at support@knowledge-architecture.com

Synthesis Search requires configuration of the following managed properties:

Title – Searchable, Queryable, Retrievable

Path – Retrievable

Size – Queryable, Retrievable, Refinable

DefaultEncodingURL – Retrievable

PictureThumbnailURL – Retrievable

ListId – Queryable, Retrievable

ListItemId – Queryable, Retrievable

LastModifiedTime – Retrievable, Refinable, Sortable

FileType – Queryable, Retrievable

contentclass – Queryable, Retrievable

URLOWSURLH – Retrievable

AuthorOWSUser – Searchable, Queryable, Retrievable

SPContentType – Queryable, Retrievable

WebId – Retrievable, Refinable

SecondaryFileExtension – Queryable, Retrievable, Refinable

StartDate – Map ows_q_DATE_EventDate crawled property to this. Queryable, Retrievable, Refinable, Sortable

owsURL – This property is custom has to be created. Map ows_URL crawled property to it. Set to Queryable, Retrievable, Safe.

Filename - Map ows_FileLeafRef crawled property to this. Move the ows_FileLeafRef field to the top and select "Map the first crawled property that contains a value". Under Advanced Search Settings, set this property to search Context 1.

 

APPENDIX B: Yes/No Field Search Methodology

In order to make search useful for fields containing Y/N as values, those fields are indexed as follows:

  • If the value is N, nothing is inserted into the index.
  • If the value is Y, the display name of the field is inserted into the index.

For example, if the value in the RegisteredArchitect column is Y for a particular employee, then the words “Registered Architect” will be inserted into the index for that employee. For custom fields, the configured display name is inserted.

 

APPENDIX C: Keyword Query Syntax for Microsoft SharePoint Searches 

Wiki Pages:

([user query string] OR [any file names that match internal link results]) AND SPContentType="Wiki Page" 

Events:

[user query string] AND SPContentType=Event

Links:

[user query string] AND SPContentType=Link AND -owsURL:[any internal sharepoint url]

Note: This query gives everything that matches in a link library that is a link to an external resource, and its results are then joined with all navigation link matches to external resources

Images:

([user query string] OR [any image file names that are extracted from the content of Wiki Pages]) AND (SecondaryFileExtension=[one of the image file extensions listed below] AND Size>[minimim allowable size])

Image File Extensions =  "jpg", "png", "gif", "tif", "jpeg", "tiff", "bmp"  

Minimum File Size = 5KB       

Videos:

[user query string] AND SecondaryFileExtension=[one of the video file extensions listed above]

Video File Extensions = "mov", "wmv", "mp4", "flv", "mpg", "avi", "m4v", "asf"

Documents:

([user query string] OR [any file names that match internal link results]) AND -SPContentType=Post AND -SPContentType=Comment AND -SPContentType=Categories AND -contentclass=STS_Site AND -contentclass=STS_Web AND -contentclass=STS_List AND -contentclass:STS_List_ AND -SPContentType="Wiki Page"  AND -SPContentType=Event AND -SPContentType=Link AND -SPContentType=Item AND -FileType=html AND -SecondaryFileExtension=aspx AND -SecondaryFileExtension=[one of the image file extensions listed above] AND -SecondaryFileExtension=[one of the video file extensions listed above]

Other Results:

(([user query string] OR [any file names that match internal link results]) AND -SPContentType=Post AND -SPContentType=Comment AND -contentclass=STS_Site AND -contentclass=STS_Web AND -contentclass=STS_List AND -contentclass:STS_List_ AND -SPContentType="Wiki Page" AND -SPContentType=Event AND -SPContentType=Link AND (FileType=html OR SecondaryFileExtension=aspx OR SPContentType=Item)) OR ((Title:[user query string]) AND -contentclass=STS_List_Posts AND -contentclass=STS_List_Comments AND -contentclass=STS_List_Categories AND (contentclass=STS_List OR contentclass:STS_List_))

APPENDIX D: SharePoint 2013 Search Result Ranking

We rely on SharePoint 2013 Search's ranking algorithm to determine the relevant content to return for user queries. There are a number of factors that SharePoint uses to calculate relevance, the core component being the "Ranking Model". This is an xml file that provides ranking instructions to the search engine. Knowledge Architecture uses the "Default Search Model," without any customization, as its ranking model.

A few things to note about the "Default Search Model" that will hopefully give you a better understanding of how results are ranked:

  • Some managed properties are ranked higher than others - for example, Title and Filename are ranked higher than Author and Body content. Generally, the simplest way to boost the ranking of an item is to include the most relevant search terms in the item's Title or Filename.
  • Some file types are ranked higher than others - Word documents are ranked higher than PDFs which are ranked higher than Images which are ranked higher than Generic List Items. In cases where SharePoint encounters a file type that doesn't match one of it's pre-defined file types, it tends to rank that item as a Generic List Item. So items with unusual file extensions may suffer in the rankings.
  • As with most search engines, SharePoint looks at search term matches within the context of the document - so if a search term is matched twice within a very long document, that document may rank lower than a shorter document where the search term is matched only once.
  • Furthermore, "Url Depth" is a factor in search rankings. So an item that is buried deep within a site collection will generally rank lower than one located closer to the top level sites.
  • Another factor in determining an item's ranking is whether or not it is linked to from pages considered to be "authoritative," or whether or not many users have previously clicked on that item in the past. These factors may be one reason why documents that do not visibly appear to be the best matched sometimes bubble to the top of the search results.

We do make a couple of adjustments to SharePoint's default managed properties in order to improve the relevance of Document results. We add the crawled property "ows_FileLeafRef" to the "Filename" managed property, since we have found that crawled property to better reflect the file display name of document list items. Furthermore, we change the "Context" of the "Filename" managed property to "Context 1". SharePoint uses these "contexts" to determine whether and how much to boost the rankings of particular properties within the scope of each ranking model. Within the "Default Search Model", for example, properties mapped to "Context 1" are given a relative contribution weight of 0.36, whereas properties mapped to "Context 2" are given a relative contribution weight of 0.15. By changing the context of the Filename property, we are roughly doubling the relative contribution weight of a document's file name, putting that property on par with the "Title" property of the document. We chose to do this because we have found that our clients are more likely to have well-populated file display names than document titles, which are often auto-generated by metadata.

Have more questions? Submit a request
Powered by Zendesk