Scenario: Increasing the quality and relevance of IBM Watson™ Engagement Advisor responses

Cognitive solutions like Watson™ Engagement Advisor enhance expertise and provide actionable information to increase client satisfaction. For example, a healthcare provider might offer its customers a Watson Engagement Advisor solution that identifies various condition management programs, resources and activities to improve their quality of life. Unlike traditional systems, cognitive solutions are not programmed to provide the right answers; they are trained with trusted information to produce accurate, useful responses.

IBM® Watson Curator optimizes business information for increased user confidence in the responses from Watson Engagement Advisor. It does so by simplifying the identification, review, and governance of the information collections that are used to inform and train cognitive solutions such as Watson Engagement Advisor. By efficiently guiding a curation team through this simplified process, Watson Curator increases the quality and relevance of information collections. As a result, Watson Engagement Advisor produces more accurate, useful, and defensible responses.

This scenario demonstrates how Watson Curator can be used to optimize and manage an information collection of key content. The Content Curator, working with Data Experts and Domain Experts, collects content which is then curated and used to build the Watson Engagement Advisor corpus. The corpus enables Customer Service Representatives to improve their productivity and offer their customers superior support.

1. In the process of creating the new collection, John, the Content Curator, defines the scope of the collection that is required to support the company's new product.
2. John works with Dieter, a IBM StoredIQ Data Expert, to review the All Data data map to understand what content is available. By using IBM StoredIQ Data Workbench, they apply filters to reduce the volume of data in order to efficiently begin their assessment. After refining the data multiple times, they take action to provide the data for curation.
3. John recruits Swetha, a Domain Expert, to curate the collected data and build a collection of targeted information.

   Because this is a new collection, John begins with the Collected phase, which contains new content that is ready to be curated. He reviews the content and begins the classification process for one or more documents. After completing curation for some of the documents in the Collected phase, John includes the documents in the curated document set of the collection. This content is placed in the Curated phase.

   For content that requires the knowledge and experience of a subject matter expert like Swetha to curate it, John starts the curation process, which moves the content to the Working phase. He then creates a task for the Domain Expert that includes instructions for performing a review and curating the content. Swetha reviews the content that is associated with the task for relevance and usefulness to the collection. She classifies some content and adds topics to it before including it in the curated document set of the collection. This content is placed in the Curated phase. Other content she excludes from the curated document set of the collection because it does not meet the collection scope and definition. This content is placed in the Excluded phase.

4. John assesses the collection for completeness and evaluates how the team is managing the content.

   He reviews the collection statistics and identifies white papers as an area where the collection needs more depth. To track resolution of this exposure, John creates a task for Dieter and includes instructions that detail the additional information that is needed for the collection. By using IBM StoredIQ Data Workbench, John and Dieter work together to identify the targeted data by adding filters that target white papers. When they review the results of the filter, they see a small set of word processing documents that can mitigate the current exposure in the collection. To meet any audit, compliance, and governance requirements, they create a report which documents the assessment and filtering that they performed. Finally, they take action on the identified documents, moving them into the collection as newly-collected content.

5. When John logs into Watson Curator, he sees a new task in his queue, signaling that the new content was added to the Collected phase of the collection and needs review.

   He opens the task and begins curating the new content. After reviewing and curating the content, John includes the documents in the curated document set of the collection. He then reviews the collection statistics again and confirms that his concern with the collection was resolved. Seeing that all existing documents that are relevant to the project are collected, classified, and enhanced, John completes the collection.

6. Preston, a Customer Service Representative, receives a call from a customer who has questions about the new product. By using Watson Engagement Advisor, Preston determines the answers to the customer's questions with a very high degree of confidence. He selects the Surrounding Information icon to better review any additional information, and he sees that the answers to the customer's questions are clearly documented. Preston easily resolves the call and is quickly on to the next call.