Home

Welcome to the xatkit-nlu-server wiki!

Packages

core

Core functionalities of the xatkit-nlu-server.

pipelines.py: includes tokenizers and stemmers to be used by the bots.

text_preprocessing.py: text preprocessing library.

training.py: methods for training the underlying models of bot contexts.

prediction.py: methods for predicting a text's intent and its parameters.

core/ner/: Named Entity Recognition library

dsl

Domain Specific Language, which includes classes for each Bot component.

dto

Data Transfer Objects, used to receive data (e.g. bot components to initialize a bot) or send data (e.g. the results of the intent prediction).

Methods for parsing DSL-to-DTO and DTO-to-DSL objects are included.

utils

Utilities library.

Available endpoints (see main.py)

/bot/new/

Creates a new Bot

Params: BotRequestDTO

/bot/{name}/initialize/

Initializes the bot (botdto_to_bot)

Params: name, BotDTO

/bot/{name}/train/

Trains the bot

Params: name, ConfigurationDTO

See training docs

/status/{name}/

Prints bot information

Returns: dict object

/bot/{name}/predict/

Given a text, the bot predicts the intent and its parameters (entities), if any

Params: name, PredictRequestDTO

Returns: PredictResultDTO

See predict docs