Protiviti / SharePoint Blog

SharePoint Blog

January 30
​SharePoint 2013 Search – Bankable results!

Search and surfacing relevant content is an enterprise wide issue. If more users understood how search worked, they would be more invested in tagging content as it is consumed by SharePoint. Just how does Search work? Well, have you ever deposited a check? SharePoint Search and bank deposits have a lot in common.

High level SharePoint Search Overview

Before we get to the bank, let’s take a look at how search works. Search is broken down into six major components: Search Administration (1), Crawl (2), Content Processing (3), Analytics Processing (4), Indexing (5) and Query Processing (6). The relationship among these components is illustrated in the chart below.

Nora 2.2.151.PNG

Working together, the components accomplish two major tasks: crawls and queries. At the process start content sources are crawled and at the process conclusion the information is available to querying users. Search Administration (1) doesn’t provide information transfer in this process, it simply manages the search components.

The crawl component (2) performs a crawl of the available content within content sources (2). Content sources might be lists, libraries or external file shares.  After the content is crawled, it is passed on to the Content Processing component (3). In this step, managed properties are mapped and crawled items and transformed into artifacts for inclusion within the search index (5). Next link and URL information is stored in the link database and then processed and forwarded to the Analytics Processing Component (4).

Analytics Processing consists of two significant analysis types: Search and Usage. Search analytics emphasize the addition of items to the search index to improve search relevance and recall. Other metrics include click distance and social tags.  Usage analytics emphasize actions users will need for search: statistical analysis of usage counts (viewed/clicked items), recommendations based on user interaction and activity ranking. Search analytics share where an item is located, while usage analytics describe overall interactions with an item.

After Analytics Processing is complete, search relevance items such as links and URLs are returned to the Content Processing Component (3). After content is received from the Content Processing Component the index component (5) writes this content to the search index. This component also receives requests for information contained in the search index and returns results sets to the Query Processing Component (6).

Rounding out the process, the Query Processing Component (6) receives and analyzes incoming search queries which improves the overall search result sets. The resulting queries are sent to the Index Component (5) which returns a set of search results for a particular query to the front end user making the request (6).

So let’s get to the bank!

In this scenario, the content is a check which you deposited into your account.

Nora 2.2.152.PNGIn this example, bank accounts are the crawled Content Source (2). Once you’ve visited the Bank Teller, who plays the role of Search Administration (1), your deposited check becomes part of the Crawled Content Source (2), or Bank Accounts, and the crawling process (2) takes place to consume your recently deposited check. In order to consume the check, the crawl service must be set up to regularly scan and consume new content.  If the crawl service runs on a daily basis, with no incremental updates, a bank customer would wait for their new deposit to be consumed by the service perhaps in an overnight job. Customers have more immediate expectations for refreshed data so it is important to set the crawl schedule in accordance with those expectations.

Once your check is part of the content source, the check content is read and processed (3). This represents the role of Content Processing. Details such as account number, amount, check number, etc. are considered to be check content. The check image is transformed into an artifact and passed along to the Index (5). You know, you might want to search for that check at a future date.

All of the check details are now processed forward to Analytics Processing (4), or for a bank, Item Processing. Links to the items are established and sent back to Content Processing (3) and items relevant for the Index are forwarded to it (5).

In the meantime, you’ve decided to check, or query, your account balance and login to your online banking to verify if the check has cleared (6). Your query reaches the Index (5) which is maintaining specific items/artifacts such as the check image. The index (5) returns its result to the query (6) and you are presented with your up to date bank balance as a search result. Once the journey of a single check through a bank account is charted, it’s easy to see how all of the components in the deposit process correlate to SharePoint Search and robust search results.

How would you deposit a check without the check number, amount, routing number or payee? This document would be refused for deposit by the teller (AKA Administrative governance policy) due to insufficient information.  If SharePoint users thought about their “document deposits” in a similar fashion, document metadata for search would have a stronger application and users would have a more compelling reason to apply it.  Then, obtaining relevant search results would be as simple as getting your accurate bank account balance.

-Nora

If you're interested in learning more about Project Management for SharePoint, consider attending our SharePoint Project Managment Symposium​ in June.  Nora will be co-presenting at the event. Hope to see you there!  

For free opportunities to earn PDU credit, consider registering for a SharePoint Roundtable​ in your community! ​​

About Nora Ten Broeck, PMP, MOS: SharePoint 2013: Nora is a SharePoint enthusiast with interests in engagement management, improved business processes and Nintex. Follow her @NoraTenBroeck

 

Quick Launch


© Protiviti 2019. All rights reserved.   |   Privacy Policy