Glossary of Terms

Terms	Definitions
Artifact	An Artifact is the standard Luma Knowledge representation of information (knowledge) that is identified as important for sharing within an organization. In Luma Knowledge an artifact is created by: Importing existing candidate knowledge (see Candidate Knowledge) Manually entering an artifact using the Curator’s Workbench from a template (Future) By API initiated by a client system used by a citizen curator or a subject matter expert At a minimum, an Artifact contains a unique id, a title, a description, and a link to the original candidate knowledge (internal or external) if it is not created in the Luma Knowledge. See Artifact Types.
Candidate Knowledge	Candidate knowledge include written articles, documents, publications, frequently asked questions (FAQ’s), service definitions, manuals, etc. or knowledge held in the minds of subject matter experts
Artifact Life Cycle	An artifact goes through separate 3 states; Draft, Published, and Retired. Draft - A Draft artifact is candidate knowledge that has been imported or entered into the Luma Knowledge. A draft artifact can be deleted (removed from the system) or after it is reviewed can be published. Reviewed (future) Approved (Future) Published – A Published artifact is a reviewed artifact that has made it available to end-users. A published artifact can be deleted, or retired or expired which removes it from end-user access or checked out which moves it to draft state Retired – A Retired artifact was removed from end-user access by a curator or the system (expired). A retired artifact can be moved to draft state or deleted. See Artifact Life-Cycle State Transitions
Artifact Types	Artifact types include: A “standalone” or simple artifact A simple Frequently Asked Questions (FAQ) which is an “FAQ” artifact with related questions and answers (QnA’ s)
End-User	The user is a member of an organization which can submit inquiries to the Knowledge Management system using a client system
Client System	A calling system is a computer system using any channel to make a request to the Knowledge Management system on behalf of a user
Meta-Data	Meta-data represents key information (terms or phrases) that describes an artifact. The meta-data is inferred during NLP parsing using parts of speech and dependency rules. Meta-data is then confirmed 1) against the vocabulary and 2) during the curation process. Meta-data is realized as Meta-data entities.
Meta-data entities	Meta-data entities provide a way to make a document machine-understandable. Meta-data enhances searches for artifacts and assists the curation process by identifying known terms. Meta-Data entities include: Topics Subjects Action Words Motivations Parent Topics Key Values
(Meta-data) Topic	The topic is the object of an artifact. It represents what the subject is acting on. Topics are products or services provided which are supported by an organization with artifacts. Note: Simple sentences may not contain a topic
(Meta-data) Subject	The subject of a sentence is the who or what in relation to an action.
(Meta-data) Action Word	The action word (predicate) is normally a verb which is an action by the subject
(Meta-data) Motivation	The motivation is a term or phrase that verbalizes why someone would want to see this artifact
(Meta-data) other	In addition to the OE entities above the following entities are “normally” found during normalization: Parent Topics Key Values
Federated Artifact	An artifact that is mastered in an external (knowledge) system.
Luma Knowledge Ontology	An ontology is a specification of entities in a domain (in this case the Luma Knowledge). It contains their types (classes or concepts), class hierarchies (taxonomies), properties, and relationships between classes. It is a structure (schema) that enables restrictions, rules, and reasoning. Example entities include an Artifact, a User, a Request, Topics, etc. This specification is used to build the knowledge graph (see Luma Knowledge Ontology)
Knowledge Graph	The knowledge graph holds the instances of (entities) Nodes, directional Relationships between Nodes and properties in Nodes and Relationships.
Workbench Knowledge Graph	The workbench knowledge graph is a curator’s view of the topic taxonomy contained in the knowledge graph. It is a subset of the full knowledge graph that contains tenant, topic (domain and topic), and artifact nodes.
Taxonomy	Taxonomy is the science or technique of classification
Topic (Artifact) Taxonomy	Topic Taxonomy is a hierarchical classification of topics and their artifact nodes. Note: Topic nodes are either categorical (for classification – operating system) or tangible (actual things – Windows or Linux)
Topic Nodes	Domains nodes are a special type of topic nodes. Domain nodes represent the top-level classification of a product, service, or policy. Domains are directly related to Sub-Tenant nodes. They are parents to topic nodes. Topics nodes represent classifications of artifacts and are arranged under domain nodes. Topics are in a parent-child relationship with each other and artifacts nodes. Topic nodes are either categorical or tangible. Categorical nodes do not have any artifacts attached to them. Tangible nodes normally have one or more artifacts attached to them. Tangible nodes are also referred to as Object Topics which is something an artifact is about.
Knowledge Store (Also, known as a Knowledge Base and Knowledge Library)	The knowledge base is a data store that holds the physical knowledge artifacts. The physical knowledge artifacts are either the Master Artifact or a local copy of a Federated Artifact.
Master Organization (Also, known as a Tenant organization.)	A master organization represents the primary business entity which may or may not control zero or more sub-organizations. A master organization can contain a hierarchy of sub-organizations.
Suborganization (Also, known as a sub-tenant)	A sub-organization represents a business entity within a master organization or another sub-organization.
Multi-Tenant System	Multi-tenancy is an architecture in which a single instance of a software application serves multiple customers. Each customer is called a tenant.
Curation	Curation is the process that builds knowledge artifacts. The process involves importing or manually creating a candidate artifact. That artifact is analyzed to determine questions and answers it may address, parsed (NLP parsing) to determine "search" meta-data, and then reviewed/modified by a curator then saved or automatically saved.
Semantic Analysis (also referred to as Natural Language Parsing (NLP))	Semantic Analysis processes employed in Knowledge Management includes: Document Understanding - is used to determine the questions and answers contained in a candidate artifact. If more than one question and answer pair are found the document can be broken into question-answer pairs (an FAQ). NLP Parsing and Normalization - is used to determine the fundamental meaning in a sentence. This includes tokenizing the sentence, determining part of speech (POS) of a token, categorizing the tokens into topics, subjects, predicates (action words), and motivations. The topics, subjects, predicates, and motivations are then normalized and labeled. The results are used as search meta-data.
Topics	Topics shall include: Domain Topics represent “Parent” Topics represent nodes in a hierarchical branch in the ontology. The hierarchy forms parent-child relationships where each descendent node provides more detail for the classification of an Object Topic. Parent Topics and their child relationships are originated from the OOB ontology. Additional, branches and sub-branch nodes are added as needed by curators for adding artifacts. An example hierarchy branch Software > Enterprise > Business > Cloud Services > AWS Object Topics represent the last node in a branch and are the primary topic of an artifact. Object topics are products or services provided and supported by an organization that has artifacts contained in the knowledge management system. As artifacts are curated, via import or manually added, the candidate object topic will be suggested. An example artifact that describes “how to start a Virtual Machine on AWS” would have an Object Topic of Virtual Machine (or VM). This object topic would be placed/associated with a branch containing the AWS parent topic.
Artifact Templates	Artifact Templates are used to define an artifact type. Templates will define required and optional fields and their types (minimum artifact fields are required). Templates allow a curator to manually create an artifact in the Luma Knowledge. Out of the box, a “simple” artifact and FAQ artifact will be provided.
Import Templates	Import templates describe fields and type of a document being imported into the Luma Knowledge. These are used where it may be desired to merge two or more fields into one to map to an artifact template.