Data Models

Data Models

Index Network uses 6 different schemas to store information on Ceramic Network:

Model	ID
Index	kjzl6hvfrbw6c6wr91bqjojw1znltqso445kevew3hiywjl1ior4fga60arj9xo
IndexItem	kjzl6hvfrbw6c66p7dxhk35uass66v2q42b2sdbaw7smitphfv60y9tux4obxu4
Embedding	kjzl6hvfrbw6c5wx4eb9mmw2su1q7y4m65wd8m887ulubbfn5iawpy6ukprq4va
WebPage	kjzl6hvfrbw6c5ehmhlme25mhizo670ro6jeu6h50m9fqnie4phkqy9vip8w8t0
DIDIndex	kjzl6hvfrbw6c7wlj08gruhq6gkatb8heg07btxzzeotumzpkftsnmxrf51l0b0
Profile	kjzl6hvfrbw6c6a1qzykoqwu191lnpci1z6kp7ww18dycm9ono1r37un63my1hd

Index

Rrepresents a collection of items around a specific context, acting as a structured way to organize and access related information. Index includes fields for a title, a signer's public key, and timestamps for creation, updates, and deletion. The inclusion of signerPublicKey and signerFunction specifies the use of decentralized access control mechanism through Lit Protocol, which enables secure access based on cryptographic conditions.

IndexItem

Establishes a relationship between an index and its nodes within a graph, essentially linking the metadata or content stored in an index to individual data points or nodes. Each IndexItem is associated with a specific Index and a Node on ComposeDB, tracked by their respective IDs, and it includes timestamps for creation, updates, and deletion. This model facilitates navigating from an abstract index to concrete items or nodes, making it a crucial component for organizing and retrieving data in a structured manner within a decentralized context.

Embedding

The Embedding model is focused on storing vector embeddings for items within an index, where each embedding represents high-dimensional data reduced to vectors for efficient similarity searches or machine learning applications. It includes a model name, the vector itself, optional context, description, category, and links to the related Index and Node. Very soon, the embedding model will facilitate the interoperability of vector embeddings to enable exchange of semantic information across various AI models.

WebPage Model

While Index supports any node on the ComposeDB graph, the WebPage model is an example for developers to explore and interact with the Index Network. It provides a simple structure for storing information about web pages, including title, favicon, URL, and content, along with timestamps for creation, updates, and deletion. This model is aimed at archiving web content in a decentralized manner, allowing for the retrieval and display of web page information within the Index Network. It facilitates the indexing and searching of web content, contributing to a decentralized web archive or search engine capability.

DIDIndex Model

The DIDIndex model is specifically designed for indexing decentralized identifiers (DIDs), providing a way to associate DIDs with specific indexes and categorizing them by type. Each DID can own or star in an index, with the controllerDID field indicating the DID that controls or owns the index.

Profile Model

The Profile model represents user profiles within the Index Network, containing personal information such as a controller DID, name, bio, and an optional avatar. This model allows for the creation of user-centric data within a decentralized framework, where the controllerDID serves as a unique identifier and access controller for the profile, emphasizing privacy and user control over personal information.