Skip to main content

Additional services

When uploading and indexing the data QAnswer also generated some additonal services. We describe them here briefly.


Once your data is uploaded you get out of the box autocompletion functionality. For example if you type on the cocktails dataset "marg" you will get:

i.e. you get suggestions for "margarita" (the cocktail) and "margaglass" (a cocktail glass). By selecting one of the suggestions you will basicaly select the concept behind it, like:

By using it you will be able to search using strings and concepts.

As described in Contextual information you can indicate properties that are used to associate images to entities. These properties will also be used to visualize images in the autosuggestion, like:

Similarly if you specify a property that is used to disambiguate entities you will get autosugestions like:

SPARQL Endpoint

When indexing data in QAnswer you get a SPARQL 1.1 complient endpoint. The only limitation is that named graphs are not supported.

You can ask SPARQL queries in the

SPARQL Endpoint

tab of the corresponding QA System. You can write this queries directly in SPARQL or write a natural language question and let QAnswer generate the SPARQL query for you.

Here you can see the generated SPARQL query for the question "Margarita ingredients". Then you can run the query to get the results.

You can also write your own SPARQL query, for example:

In this case we are asking over the cocktails dataset which are the most used ingredients.

SPARQL to user

One service that is set up is SparqlToUser, a service that converts a SPARQL query to a representation that is understandable by an end-user.

Concretly the service converts the SPARQL query:

SELECT DISTINCT ?s1 where { 
?s1 ?p1 <> .
?s1 ?p2 <> .
limit 1000

to the more user friendly representation:

type / Ingredients
/ is part of, is used by / Margarita

This service is used to display what the QA system understood:

or to facilitate the understanding of the generated interpretations: