lee/hpr-knowledge-base
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Initial commit: HPR Knowledge Base MCP Server

Browse Source 
- MCP server with stdio transport for local use
- Search episodes, transcripts, hosts, and series
- 4,511 episodes with metadata and transcripts
- Data loader with in-memory JSON storage

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
Lee Hanken
2025-10-26 10:54:13 +00:00
commit 7c8efd2228
4494 changed files with 1705541 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

159
hpr_transcripts/hpr3012.txt Normal file
View File

@@ -0,0 +1,159 @@
Episode: 3012
Title: HPR3012: Sample episode from Wikipediapodden
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr3012/hpr3012.mp3
Transcribed: 2025-10-24 15:08:34
---
This is Hacker Public Radio, Episode 3, 2012, for Tuesday, 18 February 2020.
Today's show is entitled, Sample Episode, from Wikipedia, PODEN.
It is the 170th anniversary show of Ken Farloon,
and is about 9 minutes long, and carries an explicit flag. The summer is.
An English microsoday of their Swedish language podcast about Wikipedia,
this episode of HPR is brought to you by archive.org.
Support universal access to all knowledge by heading over to archive.org forward slash donate.
Hello, this is Wikipedia, PODEN, with a special episode in English.
We're coming to you from WikiTech Storm 2019 in Amsterdam.
We are doing a couple of special episodes, both about the conference itself,
but also the fantastic people that are here and the projects that they care about.
So tune in and listen to our special episodes.
This is Wikipedia, PODEN, with a special episode from the WikiTech Storm in Amsterdam.
I'm here with Sandra for Konier, and you're on the team of the structured data on Commons.
What is your role on that team exactly?
That's a good question. I work with the GLAM team at Wikimedia Foundation,
so the team that works with galleries, libraries, archives, museums,
and GLAM Wiki collaborations and the structured data on Commons project.
I make sure that we are also thinking about GLAMs as users and GLAM Wiki volunteers,
as users of structured data on Commons.
So when I do pilot projects, little projects that involve GLAM with structured data.
And what is structured data on Commons in a nutshell?
If it's possible to tell it in a nutshell?
Well, let me do my best.
Wikimedia Commons, I think many people, when the Wikimedians will know it,
because they've uploaded images there for Wikipedia, etc.
It has always been like Wikipedia itself, a Wiki with text.
So you describe files with text.
But people have always been, the problem with that is that a text is, in many cases,
only one language. It's only in English mostly.
And probably if you're a Swedish speaker and you start searching for files in Swedish,
you will find less files than you would search in English, right?
And for a long time, the community has said,
we want Commons to become multilingual.
And since Wikidata came around, so Wikidata, our, you know,
multimedia, multilingual knowledge base, has come around.
People have been starting to think it would be good to integrate Wikidata
to help describe files on Commons, to make the multilingual among other things.
And that's basically what we're doing in structured data on Commons.
So now, actually, many things are deployed, so it's live.
When you go to a file on Commons, you can, you see a tab that says structured data.
And if you go to that tab, you can describe a file with actually multilingual data,
with data from Wikidata.
So if you have a photo of a table, for instance, you say the picked table,
and it will use the Wikidata item.
And even if someone, you know, searches Commons in Swahili,
they type the concept of table in Swahili, they will find your photo.
So that's basically what we've done.
We've added support for adding descriptions with Wikidata.
And what parts of this is already live that someone might have come across
that they don't even know of yet?
So if you have uploaded a file recently to Commons,
you will have seen that you have an extra step now in the upload wizard,
where you add structured data, you can add a caption,
which is a multilingual piece of text about that that will help to file to people to find your file.
And you can add statements, you can add something that is being depicted in a file,
and you can add other bits of information like who made it, etc.
And on the file pages that themselves also under the image,
you will see these two new tabs file information and structured data,
and that's new stuff that we've added.
So if you've used Commons, you might have seen those things.
Maybe you're a little bit hidden for everyone,
they're not super, super visible with they're there.
They're quite new since a few months.
And what are some plans for the future? What's coming?
What's coming? So in fact, structured data Commons was developed with an external grant,
so with funding that came from the Sloan Foundation, that's an American Foundation.
And that's grant period is now actually ending in the end of December.
So officially like the development framework is almost ending.
But we're still finishing up some things.
And one of the things we really want to get done,
and that will probably be worked on a bit after December still,
is going to be a query engine, just like for people who know we get data,
being able to ask the complex kind of questions to the data that you will also be able to do with Commons.
We are finishing actually searching the picked statements in multilingual ways
that you can do that. When you enter structured data,
you will notice that you cannot see and cannot add dates yet,
or you cannot add geographical coordinates yet.
And so those tweaks will still be added in the upcoming months.
That's the first things that we will still add in the upcoming period,
and that you will actually quite soon already see.
And are there some wild ideas that have come up during the work with this
that you haven't had the time to implement in this session,
but that you hope that someone will pick up in the future?
Yeah, for instance, we've only had the opportunity to do very basic functionalities,
like on one file page, you can add structured data per file.
But we already see that many community members or some community members
are developing tools to do more batch things.
We really hope that there will be some community members coming up with batch upload tools
that include structured data that is not there yet.
One thing that we've been thinking about a lot
is it might change the nature of galleries on Wikimedia Commons.
So now they are hand-created.
If you go to, I don't know, Barack Obama as a search German Commons,
you might find a gallery that is hand-created by someone.
But in the future with structured data,
you could imagine that these galleries are also automatically created.
Another thing that is definitely something that's difficult to develop
because it's super complex.
But it would be a beautiful kind of advanced search for Commons
that you don't just type as you currently do a word
and you only find things that have that words in the description.
But that you would get a search that is a bit like the Google image search
where you get suggestions and you can filter that would be super awesome to do.
But we've been exploring it a bit in the team.
What it would take to do it.
I'm just saying it's very complicated to do it.
It's not something you would develop in a month or so.
But it's definitely a longer term dream to have much better search on Commons as well.
So yeah.
And if we're thinking about impact to other projects and just searching
and finding things on Commons,
what do you see in the future for example for Wikipedia?
For Wikipedia, we think of situations that it's for people who write articles
it should become with structured data a whole lot easier to find appropriate images
for what they are writing about because of the link with Wikidata.
Under the hood actually, when you write a Wikipedia article about a topic
it is connected to a Wikidata item.
And then through that Wikidata item you can then go to Commons
and then find on Commons images that might be of already good quality
because that structured data about that image is already there.
So it might even help people who write Wikipedia.
But we also expect that it will help people who want to develop WordPress plugins
or something like that.
You just type a concept and you get the best images about that concept.
So searching for those kinds of functionalities and external tools should also be a lot easier.
Yeah. Things like that.
It should make it easier and more flexible for people to build tools
and more powerful tools and multilingual tools.
And we hope to see that in the future.
That sounds like a very interesting future.
I hope we get there soon.
Thanks for taking the time and talking to me into the Wikipedia pattern listeners.
Well, thank you and I hope you try it out.
You have just listened to one of our special episodes
from Wikitext Storm 2019 in Amsterdam.
We'll soon be back with more episodes.
You've been listening to HackerPublic Radio at HackerPublicRadio.org.
We are a community podcast network that releases shows every weekday, Monday through Friday.
Today's show, like all our shows.
We are a community podcast network that releases shows every weekday, Monday through Friday.
Today's show, like all our shows.
You've been listening to HackerPublic Radio at HackerPublicRadio.org.
We are a community podcast network that releases shows every weekday, Monday through Friday.
Today's show, like all our shows, was contributed by an HPR listener like yourself.
If you ever thought of recording a podcast and click on our contributing to find out how easy it really is.
HackerPublic Radio was founded by the digital dog pound and the Infonomicon Computer Club.
And it's part of the binary revolution at binrev.com.
If you have comments on today's show, please email the host directly.
Leave a comment on the website or record a follow-up episode yourself.
Unless otherwise stated, today's show is released on the creative comments,
attribution, share a life, 3.0 license.