This release is a pre-release and may not be stable for production use.
ovos-skill-wikipedia
Wikipedia skill for OpenVoiceOS. Adds a voice interface on top of ovos-wikipedia-plugin, which handles all Wikipedia search and retrieval.
The skill supports two answer modes:
- Explicit intent: handles utterances that name Wikipedia directly, such as "search Wikipedia for X" or "what does Wikipedia say about X". These always go to this skill.
- Common Query: handles general knowledge questions, such as "what is X" or "tell me about X", through the OVOS Common Query pipeline. The pipeline asks all registered knowledge skills for an answer and picks the one with the highest confidence score. Wikipedia competes alongside other knowledge skills such as Wolfram Alpha and WordNet.
If you name Wikipedia, this skill answers directly. If you ask a general question, this skill answers only when it has the most confident response among the competing skills.
Installation
pip install ovos-skill-wikipedia
Explicit intent utterances
These always route to this skill because they name Wikipedia explicitly:
- "Search Wikipedia for Ada Lovelace"
- "What does Wikipedia say about beans?"
- "Look up the Pembroke Welsh Corgi on Wikipedia"
- "Wiki chocolate"
- "Check wiki for Elon Musk"
Common Query utterances
These go through the pipeline — Wikipedia answers if it wins:
- "What is a black hole?"
- "Tell me about the Roman Empire"
- "Who was Marie Curie?"
Common Query pipeline
When the Common Query pipeline plugin is active, this skill competes against other knowledge skills (e.g. Wolfram Alpha, WordNet) to answer general questions. The pipeline selects the response with the highest confidence score.
Sub-plugins
All sub-plugins are optional. Configure them via ~/.config/mycroft/skills/ovos-skill-wikipedia.openvoiceos/settings.json:
{
"extractive_qa": "ovos-bm25-solver",
"reranker": "ovos-bm25-reranker",
"keyword_extractor": "ovos-rake-keywords"
}
Reranker (reranker)
Wikipedia searches often return several candidate pages for ambiguous topics. A search for "Mercury" can return pages about the planet, the element, or the Roman god. Without a reranker, the skill orders results only by a fuzzy match between the query and the page title or summary.
A reranker re-scores all candidate pages against the original query with a cross-encoder model, and moves the most relevant page to the top. This step improves voice answer quality more than any other setting, because a wrong page pick gives a completely irrelevant spoken answer.
{ "reranker": "ovos-bm25-reranker" }
For better accuracy at the cost of more CPU, use a neural cross-encoder like
ovos-flashrank-reranker-plugin.
Extractive QA (extractive_qa)
By default the skill speaks the full introductory summary of the best Wikipedia article, which can be several sentences long. An extractive QA plugin reads that summary and extracts the single passage most likely to directly answer the user's question.
Without extractive QA — "Tell me about Ada Lovelace":
"Ada Lovelace was an English mathematician and writer, chiefly known for her work on Charles Babbage's proposed mechanical general-purpose computer, the Analytical Engine. She was the first to recognise that the machine had applications beyond pure calculation..."
With extractive QA — same query:
"Ada Lovelace was the first to recognise that the machine had applications beyond pure calculation."
Much more natural for a voice interface.
{ "extractive_qa": "ovos-bm25-solver" }
For better precision, use a reading-comprehension model like
ovos-roberta-qa-plugin.
Keyword Extractor (keyword_extractor)
The initial Wikipedia search can return no results for conversational phrasing, such as "tell me more about that thing on the moon". When this happens, the keyword extractor pulls the most relevant terms from the query and retries the search with a cleaner keyword.
Without it, over-specific or colloquial queries silently return nothing. With it, the skill recovers and finds the right page in most cases.
{ "keyword_extractor": "ovos-rake-keywords" }
License
Apache 2.0 — see LICENSE.
1 maintainer
Metadata
1 maintainer
Release files for ovos-skill-wikipedia 0.18.5a1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| ovos_skill_wikipedia-0.18.5a1.tar.gz | 915.0 kB | Without provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| ovos_skill_wikipedia-0.18.5a1-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 1.9 MB
Release files / ovos_skill_wikipedia-0.18.5a1.tar.gz
| Download URL | ovos_skill_wikipedia-0.18.5a1.tar.gz |
|---|---|
| Size | 915.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
22265a4d9658e4cfdb5f4208a934bb492b6985f7cd4db7c5f3610c5b43fa225c
|
|
BLAKE2b-256 checksum How to use checksums |
5fcfdab3911a7ae3a6a670b3d85326f0b80801d05faba5e2262b9201288eb4fe
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.16
|
Release files / ovos_skill_wikipedia-0.18.5a1-py3-none-any.whl
| Download URL | ovos_skill_wikipedia-0.18.5a1-py3-none-any.whl |
|---|---|
| Size | 966.4 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
db29ef12dc8cb672f04e01a943aa4a68243e8211fa771e095fae4086161c963e
|
|
BLAKE2b-256 checksum How to use checksums |
a58b1b2bbc2d13dadc75147e0ab73a0c0c6a09cebb349f8f9e61ddaa30512972
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.16
|
Security for ovos-skill-wikipedia 0.18.5a1
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
