Some entities in OBO ontologies are documented with a database_cross_reference --> PMID as reference for an entry. For example, https://ontobee.org/ontology/CL?iri=http://purl.obolibrary.org/obo/CL_0000742 has the definition A round chondrocyte that first differentiates in the late embryonic growth plate of bone. [database_cross_reference: PMID:15951842]
I am currently using the following snippet to retrieve a table of entities in an ontology:
robot export --input data/$1.owl \
--header "ID|LABEL|IAO_0000115|hasDbXref|subClassOf [ID]" \
--export data/$1.csv
Is there a way to also retrieve these cross references linked to definitions? In the .owl file it looks like:
<owl:Axiom>
<owl:annotatedSource rdf:resource="http://purl.obolibrary.org/obo/CL_0000742"/>
<owl:annotatedProperty rdf:resource="http://purl.obolibrary.org/obo/IAO_0000115"/>
<owl:annotatedTarget rdf:datatype="http://www.w3.org/2001/XMLSchema#string">A round chondrocyte that first differentiates in the late embryonic growth plate of bone.</owl:annotatedTarget>
<oboInOwl:hasDbXref rdf:datatype="http://www.w3.org/2001/XMLSchema#string">PMID:15951842</oboInOwl:hasDbXref>
</owl:Axiom>
ROBOT doesn't support retrieval of axiom annotations using the export command, so the only way to do this in ROBOT is with a custom SPARQL query over the OWL reification model, as this is the way cross references linked to definitions are stored in the Cell Ontology (CL) and many other ontologies.
Note that other ontologies such as the Ontology of Biomedical Investigations (OBI) do this in a different way, using a plain entity annotation, this is retrievable using ROBOT export. Always remember that ontologies use annotations in a heterogeneous fashion, so approaches that work on one may not work on another (see this issue for context and discussion of unifying these styles).
If you are willing to consider a different approach, OAK can be used to retrieve definitions alongside axiom annotations, bridging across the different modeling styles in different ontologies, see this example.
Related
Are there any example ontologies where the same word has different meaning in different contexts?
For example, when building an ontology for a large company, it is not uncommon for different departments and systems to have a different definition and understanding of common words like "customer", "account", etc.
Is there a generally accepted way to model this in Protege that preserves the original words in their context, while also introducing a layer of disambiguating words for enterprise use?
This is a problem we encounter often in the biological community. I.e., the concept Eye is very dependent on the context, i.e. human eye vs fish vs spider eye etc. You can see a search for eye on the Ontology Lookup Service (OLS) and the results it return for eye from different ontologies. Disclosure: I am responsible for this tool.
Provide an IRI for your concept. This IRI should be similar to a surrogate key for your concept. I.e., instead of giving your Account concept an IRI like http://MyBusiness/someBusinessContex/Account you give it an IRI like http://MyBusiness/someBusinessContex/Context0000001. For the Eye concept the IRI for a human eye is http://purl.obolibrary.org/obo/NCIT_C12401 and for an insect it is http://purl.obolibrary.org/obo/SIBO_0000086.
I explain in this StackOverflow question the reason for using "surrogate keys".
Assign a context specific label and definition to your concept. You can use rdfs:label for label and rdfs:comment or skos:definition for definition.
You may find that you need alternatives for you concept. I.e. may be you refer to customers also as members. In this case you can use skos:altlabel to provide alternative names for your concept and skos:preflabel to define a preferred label.
So how does this work? For user interfaces you make use of rdfs:label/skos:preflabel and rdfs:comment/skos:definition for display purposes. From a data integration perspective you use the IRI.
I'm not too familiar with Machine Learning techniques, and i want to know if I can transfer a final trained-model to another machine. More specifically, i'm trying to solve a sound classification problem by training a model on a regular PC, and then implement / transfer its output model to an embedded system where no libraries are allowed (C programming). The system does not support file reading either.
So my question is.
Are there learning methods with output models simple enough that it can be implemented easily on other systems? How would you implement it? (Something like Q-learning? although Q-learning wouldn't be appropriate in my project.)
I would like some pointers, thanks in advance.
Any arbitrary "blob" of data can be converted into a C byte array and compled and linked directly with your code. A code generator is simple enough to write, but there are tools that will do that directly such a Segger Bin2C (and any number of other tools called "bin2c") or the swiss-army knife of embedded data converters SRecord.
Since SRecord can do so many things, getting it to do this one thing is less than obvious:
srec_cat mymodel.nn -binary -o model.c -C-Array model -INClude
will generate a model.c and model.h file defining a data array containing the byte content of mymodel.nn.
I am using OWLAPI for a project, and I need to compare two ontologies for differences between them. This would ignore blank nodes so that, for instance, I can determine whether the same OWL restrictions are in both ontologies. Not only do I need to know whether there are differences, but I need to find out what those differences are. does such functionality exist in the OWLAPI, oz is there a relatively simple way to do this?
The equality between anonymous class expressions is not based on the blank node ids - anonymous class expressions only have blank nodes in the textual output, in memory the ids are ignored. So checking if an axiom exists in an ontology will by default match expressions correctly for your diff.
This is not true for individuals - anonymous individuals will not be found to be the same across ontologies, and this is by specs. An anonymous individual in one ontology cannot be found in another, because the anonymous individual ids are scoped to the containing ontology.
Note: the unit tests for OWLAPI have to carry out a very similar task, to verify that an ontology can be parsed, written and parsed again without change (i.e., roundtripped between input syntax and output syntax), so there is code that you can look at to take inspiration. See TestBase.java - equal() method for more details. This includes code to deal with different ids for anonymous individuals.
I've been looking for a way to 'productionize' R or python based Random Forest/Gradient boosting tree models, and had thought that since all the individual component decision tree are binary trees, exporting to a graphical database might be a workable solution (deploying by holding the models in memory and invoking from a lightweight restful library like Flask doesn't scale that well). Here's how a decision tree is normally traversed:
1.) Data gets passed to the root node
2.) We check if the present node is a leaf node; if it is, we return a set of attributes (the predicted distribution/value).
If not, the node stores a decision rule, and checks the relevant column for which node to pass the data to next (e.g., "If age>9.5, move to left node")
Repeat 2-3.
I'm new to neo4j and graph databases in general, and it wasn't clear to me that it is possible to store(and subsequently traverse) decision rules in a node; all the examples I saw tended to be in the vein of
MATCH (neo:Database {name:"Neo4j"})
MATCH (johan:Person {name:"Johan"})
CREATE (johan)-[:FRIEND]->(:Person:Expert {name:"Max"})-[:WORKED_WITH]-> (neo)
where the conditional statements are prespecified in a query. Is this something which is feasible with neo4j, and if so, which areas of the documentation should I be focusing on?
Thank you for any guidance you could provide.
Interesting problem.
You need a way to export a model out of R or Python and translate that into a Neo4J graph.
The export mechanism can be PMML (if you're using R rpart package to generate prunded trees), Google protobuf (if you're using R gbm package to generate trees), or simply an Excel spreadsheet.
Parsing and unmarshalling to Neo4J is your issue.
I am not affiliated with Yhat in any way, but reading your question made me think of an alternative approach.
Yhat Science Ops
I don't know what that means for your team internally, but it seems like a pretty simple way to have a model easy to call via a basic API call.
i have given scene in webGL of a pizza for example. Now i want to make the content in the scene semantic visible for maschines with a custom owl and rdf for example.
When i make a owl ontologie with protege can i just use this ontology like here
https://support.google.com/webmasters/answer/146898 and instead of linking to
http://rdf.data-vocabulary.org/# i link to my own ontology and use my properties, tags etc. ?
WebGL is an API. Any data is outside the scope of WebGL. There is nothing to markup or add any ontology to.
Maybe you're looking for something like X3DOM?