Nebulento¶

In a nutshell

Nebulento is an intent parser that figures out what a user meant even when they make typos, reorder words, or phrase things loosely. It matches by fuzzy similarity rather than exact words. Think of it as a more forgiving sibling of Padatious. You still give it example sentences, but it tolerates messy input. It needs no training step and no model files. See the Glossary for terms like intent and entity.

Not in the default pipeline. Add its stage IDs to intents.pipeline explicitly to use it.

A lightweight fuzzy-matching intent parser.

nebulento finds the closest matching intent by comparing the utterance against all of an intent's training sentences using configurable fuzzy similarity strategies (built on rapidfuzz). It handles spelling errors, word-order variation, contractions, and natural phrasing that exact-match parsers would miss. It is best suited for small-to-medium intent sets (dozens to hundreds of training sentences per intent).

Install¶

pip install nebulento            # the library
pip install --pre "nebulento[ovos]>=0.5.0a1"    # + the OVOS pipeline plugin

Usage¶

from nebulento import IntentContainer

container = IntentContainer()
container.add_intent("hello", ["hello", "hi there", "hey"])
container.add_intent("weather", ["what's the weather", "weather in {place}"])

match = container.calc_intent("helo there")   # note the typo
print(match)   # best fuzzy match, with name, entities and a confidence score

In the OVOS pipeline¶

Installed as a pipeline plugin (nebulento[ovos]), Nebulento listens on the same padatious:register_intent bus events as Padatious. This makes it a drop-in alternative. Skills register their .intent files exactly as before. See Pipelines Overview for how matchers are ordered and configured.

The package registers two opm.pipeline entry points: the flat ovos-nebulento-pipeline-plugin, and a hierarchical variant, ovos-nebulento-hierarchical-pipeline-plugin. Use the flat plugin by default. Switch to the hierarchical one when you have many skills with lexically distinct vocabulary and want domain-scoped matching, or the domain_threshold off-topic gate.

Nebulento picks a match strategy (e.g. TOKEN_SET_RATIO) to control how it scores similarity, and supports context gating to narrow which intents are even considered for a given utterance. See the plugin's own docs for the full list of strategies and how to set up context gating.

When to choose Nebulento

Use it when users phrase things inconsistently or make typos (e.g. typed input, noisy STT), and your intent set is small enough that fuzzy comparison stays fast. For large intent sets or strict matching, prefer Padatious (neural) or Adapt (keyword).

Read next: Hierarchical KNN Pipeline Related: Padatious Pipeline · Padacioso · Palavreado