The interlinking of datasets published in the Linked Data Cloud is a challenging problem and a key factor for the success of the Semantic Web. Manual rule-based methods are the most effective solution for the problem, but they require skilled human data publishers going through a laborious, error prone and time-consuming process for manually describing rules mapping instances between two datasets. Thus, an automatic approach for solving this problem is more than welcome. We propose a novel interlinking method, SERIMI, for solving this problem automatically. SERIMI matches instances between a source and a target datasets, without prior knowledge of the data, domain or schema of these datasets. Experiments conducted with benchmark collections demonstrate that our approach considerably outperforms state-of-the-art automatic approaches for solving the interlinking problem on the Linked Data Cloud.
In this repository you encounter:
- the SERIMI tool.
- and the OAEI 2010 reference alignment for Dailymed-TCM that we fixed.
SERIMI ONLY works over Virtuoso Openlink (download here) Sparql Endpoints. Therefore, you need to provide to SERIMI, as target for the interlinking, a Virtuoso Sparql endpoint.
How To Load RDF Data into Virtuoso?
To try SERIMI over you own data, you need to load your data into a Virtuoso server. Below we show how to do this.
This example assumes that you want to load Geonames RDF dataset into Virtuoso.
Notice that Virtuoso does not have a unique repository for each dataset. It has just a repository and it organizes the datasets in differents NAMED GRAPHS. The example below loads the data into the https://geonames.org Named Graph.
ALL COMMANDS ARE COMPULSORY
checkpoint_interval(-1);
DB.DBA.RDF_OBJ_FT_RULE_ADD (null, null, 'index_local');
DB.DBA.VT_INC_INDEX_DB_DBA_RDF_OBJ ();
CREATE BITMAP index RDF_QUAD_POGS on DB.DBA.RDF_QUAD (P,O,G,S);
CREATE BITMAP index RDF_QUAD_PSOG on DB.DBA.RDF_QUAD (P,S,O,G);
CREATE BITMAP index RDF_QUAD_SOPG on DB.DBA.RDF_QUAD (S,O,P,G);
//For RDF format
DB.DBA.RDF_LOAD_RDFXML_MT (file_to_string_output ('/tmp/all-geonames.rdf'), '', 'https://geonames.org');
//For NT format
DB.DBA.TTLP_MT (file_to_string_output ('/tmp/all-geonames.ttl'), '','https://geonames.org');
checkpoint;
checkpoint_interval(30);
SERIMI is a Ruby application and demands the JRuby version of ruby. We recommend to install JRuby using the RVM.
You must use any Jruby with Ruby 1.8.7. It does not work for ruby > 1.8.7.
After install JRuby, you have to install the following gems:
- actionmailer (2.3.2)
- actionpack (2.3.2)
- activerecord (2.3.2)
- activeresource (2.3.2)
- activesupport (2.3.2)
- amatch (0.2.5)
- bouncy-castle-java (1.5.0146.1)
- elasticsearch (0.0.0)
- jruby-launcher (1.0.7 java)
- jruby-openssl (0.7.4)
- json (1.5.1 java)
- OptionParser (0.5.1)
- patron (0.4.9)
- rails (2.3.2)
- rake (0.8.7)
- sources (0.0.1)
- Text (1.1.2)
- text (0.2.0)
- uuidtools (1.0.7)
- xml-object (0.9.93)
- yajl-ruby (0.8.2)
IMPORTANT: You MUST install the uuidtools version 1.0.7
gem install uuidtools -v=1.0.7
gem install xml-object -v=0.9.93
gem install rails -v=2.3.2
You can download the source code of SERIMI using the Git command below. For that, you need to install the Git version control in your computer.
git checkout git:https://github.com/samuraraujo/SERIMI-RDF-Interlinking.git
You can also download SERIMI by clicking in the button DOWNLOADS on the top of this page.
Go to the root of the SERIMI directory and executed the command below:
ruby serimi.rb
If everything is fine, this command will print help information about SERIMI. Below you find information about how to use SERIMI.
Usage: serimi.rb [parameters]
Example of use: The following example, shows how you can interlink the class StatesAndTerritoriesOfIndia from Yago with DBPedia.
ruby serimi.rb -s https://dbpedia.org/sparql?default-graph-uri=https://dbpedia.org -t https://dbpedia.org/sparql?default-graph-uri=https://dbpedia.org -c https://dbpedia.org/class/yago/StatesAndTerritoriesOfIndia
In the example above, the source is the Sider endpoint, the target is the DBPedia endpoint , and the source class to be interlinked is the class Drugs. Notice that the performance of the Sider and DBPedia endpoints may delay the output of SERIMI.
Parameters
-v, --verbose Output more information
-s, --source URI (MANDATORY) Source Virtuoso sparql endpoint - URI
-t, --target URI (MANDATORY) Target Virtuoso sparql endpoint - URI
-c, --class URI (MANDATORY) Source class for interlink - URI
-o, --output FILE Write output to FILE - Default=./output.txt
-a, --append-output value Append output to FILE - A value: a or w - Default=w
-f, --output-format value Output format: txt, nt. Default=txt
-k, --chunk value Number of source instances processed per interaction, a value >= 2 - Default=20
-p, --top k results Return only Top K results, a value >= 1 - Default=0
-b, --offset value Start processing from a specific offset - Default=0
-x, --string-threshold value String distance threshold. A value between (0,1) - Default=0.7
-y, --rds-threshold value RDS threshold. A value between (0,1) - Default=max(media,mean)
-u, --use-pivot value Select a pivot to reinvorce the class of interest. A value (false or true) - Default=false
-m, --sort-source value Sort resources before appling the selection phase. A value (false or true) - Default=true
-l, --logfile FILE Write log to FILE
-h, --help Display this screen
You can change the value of thresholds used in SERIMI. There are three parameters for this:
-x: allow you to define a threshold for the string distance function applied in SERIMI. It selects labels of the sources resources and search for these labels in the target endpoint. This parameter defines how similar the retrieved resources should be with regards to the searched label. You can define a value between (0,1)
-y: allow you to define a threshold for the RDS function implemented in SERIMI. Currently the value for this parameter is computed automatically. You can force a specific value for this parameter. You can define a value between (0,1).
-k: allow you to the define the number of instances to be interlinked per interaction. Different values may affect the precision and recall. Small values work better for datasets with a lot of instances sharing a same label. Ex. DBPedia.
SERIMI outputs the interlinks in an external file (output.txt is the default). It accepts two output formats (text or nt). You can configure the output format using -f parameter.
For instance, given the command below:
ruby serimi.rb -s https://www4.wiwiss.fu-berlin.de/sider/sparql -t https://dbpedia.org/sparql?default-graph-uri=https://dbpedia.org -c https://www4.wiwiss.fu-berlin.de/sider/resource/sider/drugs -f nt
The output would contain the triples:
<https://www4.wiwiss.fu-berlin.de/sider/resource/drugs/1046> <https://www.w3.org/2002/07/owl#sameAs> <https://dbpedia.org/resource/Pyrazinamide>
<https://www4.wiwiss.fu-berlin.de/sider/resource/drugs/104758> <https://www.w3.org/2002/07/owl#sameAs> <https://dbpedia.org/resource/Raltitrexed>
<https://www4.wiwiss.fu-berlin.de/sider/resource/drugs/10631> <https://www.w3.org/2002/07/owl#sameAs> <https://dbpedia.org/resource/Medroxyprogesterone_17-acetate>
<https://www4.wiwiss.fu-berlin.de/sider/resource/drugs/1065> <https://www.w3.org/2002/07/owl#sameAs> <https://dbpedia.org/resource/Quinidine>.
......
......
......
To report any issue about this tool you can use this system: https://github.com/samuraraujo/SERIMI-RDF-Interlinking/issues
[Here] (https://github.com/samuraraujo/SERIMI-RDF-Interlinking/tree/master/SERIMI-TECH-REPORT-v2.pdf) you can find a technical report about SERIMI.
SERIMI is distributed under the LGPL[https://www.gnu.org/licenses/lgpl.html] license.
SERIMI 0.2 released. (git master)
We participate in the OAEI 2011. Below you find the list of parameters used in our evaluations:
-k 10 -p 1
We use the same parameters in all pairs of datasets compared.