Preloaded RDF databases

To help users get started with RDFGraphs, we’ve uploaded ready-made databases for a few examples of popular RDF datasets that have already been imported to Kùzu. You can download these databases without any additional setup and start querying them in Cypher with Kùzu.

Note on database versions: The below databases have been preloaded to work with Kùzu version 0.2.0. Because Kùzu’s storage continually evolves, it’s important to use the same version of Kùzu as was used to create the database, in this case, v0.2.0.

Tip

In general, you can easily share your Kùzu databases with others by zipping the database directory and send them the zip file via conventional means like object stores or shared drives. All you have to do is to ensure that both parties use the same Kùzu version.

Preloaded RDFGraphs

The names shown in the first column below are the RDFGraph names the RDFGraphs in Kùzu you will see when you
download these databases.

Info

Scroll to the right to see the full table.

Download URL	Description	Database Size zipped/uncompressed	# Resources	# Literals	# Triples	License	Details
DBPedia	Entire DBPedia knowledge graph. See: dbpedia.org	25GB/100GB	129037240	344778409	867469010	CC 3.0 and GNU Free Doc. See here.	Converted from latest core DBPedia dump on Feb 15-2024 from here.
EWordnet	English Wordnet lexical database. See: WordNet	68MB/367MB	987816	494916	2892034	CC 4.0	Converted from this english-wordnet-2023.ttl.gz file.
Geonames	Popular geographical database. See: geonames.org	3.7GB/20GB	49213751	63882790	181846462	CC 4.0	Converted from this all-geonames-rdf.zip file.
OpenCyc	Popular common sense knowledge graph. See: Cyc project.	70MB/359MB	608573	1083192	2413894	CC	Converted from this opencyc-latest.owl.gz file. Skipped 43755 triples with malformed integers.
WKLexemes	Wikidata lexicographical information. See: here	2.5GB/15GB	23477554	42780808	162821310	CC 3.0	Converted from this wikidata-20240209-lexemes-BETA.ttl.gz file.
YagoSchema	Upper taxonomy, constraints, & property definitions of Yago. See: here	81KB/16MB	668	48	1085	CC 3.0	Converted from the yago-schema.ttl file extracted from this yago-4.5.0.2.zip file.
YagoTaxonomy	Full taxonomy of Yago 4.5. See: here	7.7MB/50MB	132883	0	166366	CC 3.0	Converted from the yago-taxonomy.ttl file extracted from this yago-4.5.0.2.zip file.
YagoFacts	Facts about entities that have an English Wikipedia page in Yago 4.5. See: here	4.7GB/31GB	11138914	284724159	312652091	CC 3.0	Converted from the yago-facts.ttl file extracted from this yago-4.5.0.2.zip file.
YagoBWk	Facts about entities that do not have an English Wikipedia page in Yago 4.5. See: here	20GB/136GB	89522405	1322909879	1472516819	CC 3.0	Converted from the yago-beyond-wikipedia.ttl file extracted from this yago-4.5.0.2.zip file.
Yago	Entire Yago 4.5 knowledge graph. See: here	25GB/166GB	99313897	1607634086	1785336361	CC 3.0	Union of the above 4 Yago files.

To verify your download, please use the MD5 checksums here.

If you would like to see another popular RDF dataset in Kùzu hosted here, please email us or chat with our team on Discord.

Getting Started With a Preloaded Kùzu RDFGraph

Below are the instructions for the English Wordnet that is stored as EWordnet RDFGraph. The same instructions apply to all other RDFGraphs.

Download the zip file

Pick a location to download the Kùzu database from. Let ${KZ-DB-BASE-DIR} be the address of that location. The commands for doing this from your terminal are showb below, but you also can download it to ${KZ-DB-BASE-DIR} directory by just clicking on the above link and double clicking on the zip file.

> cd ${KZ-DB-BASE-DIR}
> wget https://rgw.cs.uwaterloo.ca/kuzu-rdf-database/0.2.0/zips/e-wordnet-kz.zip
> unzip e-wordnet-kz.zip
> ls
e-wordnet

If the download was successful, you should see the e-wordnet directory when you run the ls command in ${KZ-DB-BASE-DIR}.

Now you are ready to query and explore the database in Kùzu via Cypher! Below, we show how you query the database with Kùzu CLI but you can also use Kùzu Explorer or any of Kùzu’s client libraries.

Visualize the data in KùzuExplorer

To get familiar with the database and its schema, you can use KùzuExplorer, our visualization tool. See the download instructions here (note that you need Docker installed and running on your machine first).

Connect the Docker server to your local directory volume by specifying the -v flag and the absolute path to the database directory as follows:

docker run -p 8000:8000 \
    -v /path/to/your/local/database:/database \
    --rm kuzudb/explorer:latest

You can visualize the EWordnet RDFGraph schema from the Explorer by clicking on the “Schema” tab on the top right.

Run the following query in the editor to see 20 random triples as follows:

MATCH (s:EWordnet_r)-[p:EWordnet]->(o)
RETURN * LIMIT 20;

Query the database using the CLI

If you prefer to work in the terminal, you can just as well use the Kùzu CLI to query the database.

Follow the instructions here and replace the HTTPS URL to download the CLI compatible with Kùzu version 0.2.0.

Connect to the downloaded database’s directory via the terminal as follows:

./kuzu ${KZ-DB-BASE-DIR}/e-wordnet

Then run the following command on your shell to start the CLI and issue a query:

Display the available tables in the database as follows:

kuzu> CALL show_tables() return *;

Output:

------------------------------------
| name        | type     | comment |
------------------------------------
| EWordnet_r  | NODE     |         |
------------------------------------
| EWordnet_l  | NODE     |         |
------------------------------------
| EWordnet_rt | REL      |         |
------------------------------------
| EWordnet_lt | REL      |         |
------------------------------------
| EWordnet    | RDFGraph |         |
------------------------------------
(5 tuples)

You can also run a count query to see the number of triples that match a specific condition:

kuzu> MATCH (s:EWordnet_r)-[p:EWordnet]->(o)
      RETURN COUNT(*);

Output:

----------------
| COUNT_STAR() |
----------------
| 2892034      |
----------------