Due to recent updates, all users are required to create an Altair One account to login to the RapidMiner community. Click the Register button to create your account using the same email that you have previously used to login to the RapidMiner community. This will ensure that any previously created content will be synced to your Altair One account. Once you login, you will be asked to provide a username that identifies you to other Community users. Email us at Community with questions.
Recommender: complex inference mechanism
Learner III
Hi rapid-i community,
I like to build a recommendation system using an ontologie. Thanks to the RMonto there's no problem to use RDF data. Further, my aim is to use a complex inference mechanism which analyses the properties paths (related instances) of already by the individual user rated items to find the most-likely matching ones. The mechanism is called 'filtering by property sequence', calculates the 'semantic intensity' of each item along the sequence of related items to the user rated item(s) and stops if the semantic intensity fall below a certain threshold. (The benefit of this approach is that unknow relations can be identifiyed und interpreted in the context of the (known)user's preferences.)
I'd like to ask, if anybody has an idea how to model this process in RM.
So far my only approach was to to make SPARQLs, but this seems to miss some possible property sequences and is also not the most efficient one. Unluckyly the search for an extension or at least a code basis for java plugin was without results. So, it would be great if there's someone with RDF/ OWL expertise who has a smart idea!
Thanks in advance.
Heinz
I like to build a recommendation system using an ontologie. Thanks to the RMonto there's no problem to use RDF data. Further, my aim is to use a complex inference mechanism which analyses the properties paths (related instances) of already by the individual user rated items to find the most-likely matching ones. The mechanism is called 'filtering by property sequence', calculates the 'semantic intensity' of each item along the sequence of related items to the user rated item(s) and stops if the semantic intensity fall below a certain threshold. (The benefit of this approach is that unknow relations can be identifiyed und interpreted in the context of the (known)user's preferences.)
I'd like to ask, if anybody has an idea how to model this process in RM.
So far my only approach was to to make SPARQLs, but this seems to miss some possible property sequences and is also not the most efficient one. Unluckyly the search for an extension or at least a code basis for java plugin was without results. So, it would be great if there's someone with RDF/ OWL expertise who has a smart idea!
Thanks in advance.
Heinz
0
