Best Practices for Knowledge Building

A Knowledge Artifact is the source of knowledge shared within the organization or with the external world. It is the basic building block of any Knowledge Management System. Artifacts can be created from a document, web source, multimedia file, or any other content that contains information deemed important for knowledge sharing. Well-written Knowledge articles are easy to search and read, which enables Support analysts and End-users to diagnose and resolve their issues effectively. Poorly constructed artifacts may not be processed correctly and therefore may not provide relevant information for user inquiries.

Here are some best practices that should be followed when creating Knowledge Artifacts in Luma Knowledge:

1. Ensure the Document is well Structured
   Document or Webpages used to create should be structured appropriately. Luma Knowledge generates QnA pairs from the uploaded document contents. For successful Artifact creation, the system expects the documents in form of Questions and Answers (Heading-Content pairs). Ensure that the source document has at least one heading.

2. Identify appropriate Topic Hierarchy, before creating the Artifact
   Topic linked to a Knowledge Artifact indicates the product or service, the artifact is linked to. Artifact in Luma Knowledge should be linked to the correct topic hierarchy so that the knowledge appears in relevant searches. Before creating the artifact, locate or create the required Topic in the Knowledge Graph.

3. Well-formed artifact Summary
   Luma Knowledge uses Summary to identify the metadata for an Artifact. The text in the Summary field is used by the NLP engine to generate Metadata information. It provides a reference to the Knowledge available in the artifact. This metadata is used to find and present the most suitable information for the user inquiry. It is important that the Summary is well-formed and articulates the information in the document well.

4. Correct Metadata tagging
   Metadata linked to the Knowledge Artifacts are used to identify the information available in the Artifact. It ensures that the artifact is searchable and provides end-users with relevant information. Ensure correct metadata is identified for the artifact. Always verify and update the metadata to represent the information available in the artifact.

5. Do not use the same metadata definition for multiple Artifact
   Luma Knowledge identifies a knowledge artifact based on the metadata linked to it. Multiple artifacts with the same metadata set indicates that the information in the artifacts may also be duplicated. This also leads to inaccurate results for a search query. Ensure that the metadata for the artifact is appropriately tagged and refers to the exact information available in the article.

6. If the Source document is not structured as Question-Answers, use ‘Import Document with Template’ option
   Not all documents in an organization follow the Question-Answer structure. When creating artifacts, Luma Knowledge breaks a document into smaller artifacts and generates QnA pairs. Such a document may get rejected during the Semantic Analysis. Use the ‘Document with Template’ option to import the document to create the artifact. This by-passes the FAQ generation process and uses the document content to create the artifact.

7. Add Effective Vocabulary
   Luma Knowledge uses the vocabulary to understand the organizations language and interpret user requests. Ensure an effective vocabulary is added to the system.