How Auto Tuning Works and Its Features

Auto Tuning is machine learning applied to search results. Its most visible impact is the displaying of rich snippets in response to how-to queries. Auto Tuning has three other features: Auto Boosting, Enable Auto Spell Correcter, and Facet Interpreter. This article touches on them all. But at first, an overview of how auto tuning works.

How Auto Tuning Works

Auto Tuning keeps a record of user activity and analyzes the data thus collected to identify patterns. The patterns are then used to change the order of search results for individual users based on their activity.

The diagram captures user journey. User 1 creates four sessions whose data is stored and analyzed in SearchUnify's machine learning engine. When User 1 returns for a fifth session, she finds search results more suitable to her requirements. Then in a sixth session, the activities of the first five sessions are considered. And it goes on and on.

The activities' of User 1 have no impact on the results of User 2 unless both happen to have similar search journeys. Search results of User 2 are personalized based on the data processed from his own activity.

Features

Auto Boosting

Auto Boosting personalizes results. The personalization is based on user behavior analysis.

Let's assume that your organization has four content sources A, B, C, and D. One of your users for some reason becomes partial to consulting documents from B. If Auto Boosting is turned on, then SearchUnify will respond to this change in user behavior by presenting more documents from B in search results. The idea is to facilitate search by anticipating user needs.

  1. Select a search client where auto boosting will be applied.

  2. Toggle Enable Auto Boosting to turn it on.

Enable Auto Spell Correcter

Enable Auto Spell Corrector expands the Did You Mean? dictionary. The new words are gathered from analyzing misspelled keywords and acquiring the specialized terminology stored in your content sources.

  1. Select a search client.

  2. Toggle Enable Auto Spell Correcter to turn it on.

Misspelled Words

A search for "moniter" on most search clients will fetch zero results. But if a user, on realizing his or her mistake, looks up "monitor" next then SearchUnify can infer that "moniter" is "monitor" misspelled. Once a deduction has been made, SearchUnify will automatically correct each instance of "moniter" in subsequent sessions.

Specialized Terminology

By importing the specialized terms stored in your content sources, you can train your search client to recognize customer and product names and your business jargon. Importing is quick.

  1. Click to launch a window where you can select content sources.

  2. Select the content source, content type, and content fields and click Add.

Enable Facet Interpreter

Enable Facet Interpreter provides faceted search. When a keyword matches (or nearly matches) a facet, then that facet is applied to narrow the scope of search and provide high relevance results.

Using Enable Facet Interpreter is safe to use because it does not mar user experience in any way. Instead of replacing regular results with faceted results, it provides as many as five extra faceted results.

 

  1. Select a search client.

  2. Toggle Enable Facet Interpreter to turn it on.

You can expand your facets database with content fields. If you want to use authors as facets and all author names are stored in a content field, then use to open a window and specify that content field.

  1. Click to launch a window where you can select content sources.

  2. Select the content source, content type, and content fields and click Add.

 

Enable Rich Snippets

Enable Rich Snippets provides answers to how-to types of questions right on a search results page.

For search queries such as "How to Install SearchUnify in Salesforce?", a search client will list the top steps right on the search results page if Enable Rich Snippets is turned on.

  1. Select a search client where rich snippets will work.

  2. Toggle Enable Rich Snippets to turn it on.

Last updatedTuesday, June 23, 2020