Glossary of terms
Common terms used through-out the Searchcraft documentation and Vektron dashboard.
- Access Key
- Application
- Click Position
- Conversion
- Default Search Field
- Document
- Field
- Fuel
- Index
- Indexing
- Ingestion
- Module
- Organization
- Pilot
- Ranking
- Relevance
- Retention
- Schema
- Space G.O.A.T
- Stemming
- Stopword
- Synonym
- Vektron
- Weighting Multiplier
Access Key
An access key is a token that is used to authenticate a request to the Searchcraft API. It is used to authorize requests to the API and to control access to the data in the index.
Application
An application is a collection of indexes, access keys and users that are managed together. It could represent a website or a mobile app and may contain one or severall indexes. in a multi-tiered environment it is typical for an application to contain multiple indexes representing the data set for each environment.
Click Position
The click position is the position of a search result within a list of results when an application user clicked on that result. This information is measured and reported within Vektron.
Conversion
That measured rate at which application users click on a search result.
Default Search Field
A field on a schema that is configured to be used for a search query when a specific no specific field criteria is supplied.
Document
An individual piece of content. Similar to a record in a database. Documents in Searchcraft should a single level of depth (no nested fields) but fields may contain either a single or multiple values. Once a document is ingested it is considered immutable.
Field
A field is a defined attribute within a schema that describes a specific type of data that can be stored or indexed. It specifies the structure, format, and behavior of the data. Refer to the Field Types section for more details of the different field types offered by Searchcraft. Fields may contain either a single value or an array of values.
Fuel
The fuel gauge in Vektron that displays the current usage of the index.
Index
An index is a collection of documents that may be searched together. Conceptually it is similar to the idea of a “database” but the technical underpinnings differ. Since a search index is typically an abstraction of a primary data source, it can either represent the data structure of a single source content type or document constructed from the common fields of multiple content type sources. However, it is more common for an index to be comprised of documents from a single content type. Documents within an index should contain fields that are used for search functionality as well as search result display. The documents do not need to contain every field from the source dataset, just the fields that are used for search and results display.
Indexing
Indexing is the process that involves storing and parsing documents that makes data searchable. See Ingestion.
Ingestion
Content ingestion is the process of adding documents to an index. Documents are added to an index by sending a request to the Searchcraft API. The request contains the document data and the index to which the document should be added. The document data in in JSON format and contains the fields that are defined in the index schema. There are three supported methods of ingestion; pull, push or direct file upload in Vektron. Push ingestion is configured through one of the available integrations, via a direct API request or via the Space G.O.A.T tool. Pull ingestion is configured in Vektron and involves providing a source feed url and configuring the ingestion frequency.
Module
Modules are an a la carte bundle of additional requests that can be added to a plan to accomodate for request overages. Modules are in addition to the base plan and get triggered when the usage limit is reached.
Organization
When you sign up for a Searchcraft account, you are automatically assigned to an organization. An organization is a group of pilots that share a common billing account.
Pilot
Searchcraft Vektron users are referred to as pilots.
Ranking
The importance of a field’s value in determining the relevance of a document for a search query.
Relevance
Relevance refers to how well a document (or search result) matches a user’s query. It is the measure of a document’s usefulness or appropriateness for satisfying the search intent. Relevance determines the ranking order of documents in the search results, with the most relevant results appearing first.
Retention
Returning users that use the search function again on an application.
Schema
An index schema defines the structure of the documents within an index. It defines the fields that are available for search and display and configures the behavior and data types of those fields.
Space G.O.A.T
Space G.O.A.T (Global Object Aggregation Tool) generates and sends Searchcraft documents for ingestion given a database connection string and table name. It is a helpful option for quicking exporting from a primary data source into a Searchcraft index for situations where either a direction integration or a public feed is not available. It a push ingestion(#ingestion) mechanism. It is available as a command line tool that is downloadable from the Vektron dashboard’s tools page.
Stemming
Stemming is the process of reducing a word to its root form. This ensures variants of a word match during a search. Enabling stemming can improve query matches. Stemming is language specific. Additional details on stemming.
Stopword
A stopword is a commonly used word that is often ignored when performing a search. These words occur frequently and typically do not contribute meaningful context to a search query so they are often removed to improve search relevance. Searchcraft comes with multiple, language-specific pre-built dictionaries that pilots may utilize or you can create and manage your own. If you are a Searchcraft Cloud customer, these are managed through the Vektron UI. For self-hosted customers, these are managed via the API.
English Examples of Stop Words
- Articles: the, a, an
- Conjunctions: and, or, but
- Prepositions: in, on, at, to, with
- Pronouns: he, she, it, they
Synonym
If you have mutliple words that mean the same thing, a synonym may be defined to create a relationship between them. It is common for acronyms to be defined as synonyms, for example “New York City” and “NYC” may be defined as synonyms and would mean that a match on either would return the same results. The words “dog” and “puppy” or “cat” and “kitten” are another example where synonyms may be desired. Synonyms are helpful when dealing with industry specific terminology. If you are a Searchcraft Cloud customer, these are managed through the Vektron UI. For self-hosted customers, these are managed via the API.
Vektron
Vektron is the dashboard for managing Searchcraft pilot accounts, applications, analytics reporting, billing, indexes and access keys. This is your central hub for managing your Searchcraft experience.
Weighting Multiplier
Another term for ranking, this is the underlying multiplier value that is used by Searchcraft to determine the importance of a field’s value in determining the relevance of a document for a search query.