LarKC WP4+WP6 Teleconference on Reasoning with Noisy Data, 11.00 - 12.00, Oct. 05, 2010
The skype ID of the chair: tonganqn (Gaston's)
Participants
Gaston (VUA), Stanley (Saltlux)
Discussions
Stanley: Cases 4 and 5 can now be solved by changing the SPARQL query.
- Case 4 was solved by adding sameAs constraints to enforce the relationship sameAs(Node2,Node3).
- Case 5 is solved by changing the where clause in the SPARQL query and finishing condition.
- A junction node is the node which several next links have and retrieved by querying of next link. The query of next link query in case 4 and 5 is following.
PREFIX rsm: <http://www.saltlux.com/rsm#>
select distinct
?nextLinkID ?nextStartNode ?nextEndNode ?nlAngle
where {
?currentLink rsm:nextLink ?nextLinkID .
?nextLinkID rsm:startNode ?nextStartNode .
?nextLinkID rsm:endNode ?nextEndNode .
?nextLinkID rsm:angle ?nlAngle .
FILTER ( ?currentLink != ?nextLinkID )
}- There is two queries for next link.
- First query is the query for junction node which made by owl:sameAs.
- Next links of junction of sameAs is retrieved as following. previous query do not retrieve nodes of sameAs .
PREFIX rsm: <http://www.saltlux.com/rsm#>
select distinct
?nextLinkID ?nextStartNode ?nextEndNode
where
{
currentLink rsm:startNode ?currentStartNode .
currentLink rsm:endNode ?nextStartNode .
?nextStartNode owl:sameAs ?sameNextStartNode .
?nextLinkID rsm:startNode ?sameNextStartNode .
?nextLinkID rsm:endNode ?nextEndNode .
filter( ?currentStartNode != ?nextEndNode )
}- Second query is the query for next link while nextLink instance is not instantiated in rsm:nextLink.
- Dataset of start node and end node is defined by rsm:link instance that is essential.
PREFIX rsm: <http://www.saltlux.com/rsm#>
select distinct
?nextLinkID ?nextStartNode ?nextEndNode
where
{
currentLink rsm:startNode ?currentStartNode .
currentLink rsm:endNode ?nextStartNode .
?nextLinkID rsm:startNode ?nextStartNode .
?nextLinkID rsm:endNode ?nextEndNode .
filter( ?currentStartNode != ?nextEndNode )
}
Gaston: It could be possible to consider case 3 as a decision problem and solve it using cognitive heuristics. The problem could be phrased as deciding whether to use (the data of) road 1 or 2 to solve a given query. For some queries the road 1 would make more sense whereas for other queries road 2 would be better. The decision on whether to use road 1 or 2 could be based on the type of information that describe both roads.