The FuzzyDL-Learner System (see also latest changes)
The Project started in collaboration with Francesca A. Lisi.
FuzzyDL-Learner is an automatic learning system for OLW 2 ontologies.
|1,000||Hostel subclass of Good_Hotel|
|1,000||hasPrice_veryhigh subclass of Good_Hotel|
|0,569||hasPrice_high subclass of Good_Hotel|
|0,286||hasAmenity some (24h_Reception) and hasAmenity some (Disabled_Facilities) and hasPrice_low subclass of Good_Hotel|
|0,282||hasAmenity some (Babysitting) and hasRank some (Rank) and hasPrice_fair subclass of Good_Hotel|
|0,200||Hotel_1_Star subclass of Good_Hotel|
Here, the range of the real valued price attribute has been automatically fuzzyfied as illustrated below.
See the here for a list of ontologies on which FOIL-DL has been tested.
So far, the system supports the following algorithms
- Inspired by the well-known FOIL algorithm
- Like FOIL-DL, though a novel ensemble learning measure has been incorporated (see here)
- Based on Real AdaBoost, where a genetic algorith has been used to generate GCI candidates
- Like FOIL-DL, though a genetic algorith has been used to generate GCI candidates
- Like gFOIL-DL, but ensemble learning measure of pFOIL-DL has been used instead