NLUDB React Hooks
This library provides a set of React Hooks that makes it insanely easy to incorporate NLP into your React application.
These hooks are optimized for in-app usage. For the full set of NLUDB client operations, see the @nludb/client
Javascript library.
useNLUDB
The useNLUDB
hook provides a React-friendly way to instantiate an NLUDB client.
import {useNLUDB} from '@nludb/react-hooks';
const [nludb, nludbError] = useNLUDB({
apiKey:apiKey
})
The nludb
variable represents a client instance from the @nludb/client
library.
useEmbeddingIndex
The useEmbeddingIndex
hook provides a React-friendly way connect to an Embedding Index for semantic search.
import { useEmbeddingIndex } from '@nludb/react-hooks';
import { EmbeddingModel } from '@nludb/client'
const [{results, isSearching, error}, {reset, search, insert}] = useEmbeddingIndex({
nludb: nludb, // The client from above
name: "Index Name",
model: EmbeddingModel.QA, // The embedding model to use
upsert: true, // Load a pre-existing index
})
Search the with:
search({query: "My query"})
Reset your search with:
reset()
Use the search results provided in the isSearching
and results
variables.
useParser
The useParser
hook provides a React-friendly way connect to an Parser for text segmentation, parsing, and named entity recognition.
import { useParser } from '@nludb/react-hooks';
import { ParsingModel } from '@nludb/client'
const [{results, isParsing, error}, {reset, parse}] = useParser({
nludb: nludb, // The client from above
model: ParsingModel.EN_DEFAULT, // Default english model
includeTokens: false, // Don't return token-level data
includeEntities: true // Do return entity data
})
Parse text with:
parse({docs: ["My document"]})
Reset your parse with:
reset()
Use the search results provided in the isParsing
and results
variables.
References
This project was adapted from the following resources on the web: