TARGETING CLIENTS THROUGH KNOWLEDGE DISCOVERY FROM DATABASES

Velike podatkovne baze podjetij vsebujejo veliko znanja, ki je strokovnim službam dobro znano, pa tudi veliko novega znanja, ki ga strokovne službe ne poznajo. Z metodami umetne inteligence, kot sta metoda pridobivanja znanja iz podatkov (KDD) in metoda podatkovnega rudarjenja (DM), smo preverili možnosti odkrivanja novih znanj in uporabnost le-teh pri ciljno usmerjenem trženju v zavarovalništvu. Zanimalo nas je ali lahko z novo pridobljenim znanjem stranki ponudimo storitve, ki so bolj prilagojene njenim individualnim potrebam. Rezultati so potrdili, da poleg znanih pravil iz podatkov pridobimo tudi marsikatero novo, presenetljivo pravilo. Na osnovi takšnih novo odkritih pravil pa lahko stranko obravnavamo drugače kot smo jo do sedaj in se s svojo ponudbo storitev bolj približamo njenim potrebam. Individualna obravnava posledično povečuje zadovoljstvo strank. Z uporabo metod umetne inteligence torej lahko povečamo učinkovitost in zmanjšamo stroške poslovanja.Large company databases can offer a lot of information to technical services, but some information still remains hidden. By using methods based on artificial intelligence, such as Knowledge Discovery from Databases (KDD) and Data Mining (DM), we tried to find new and useful information which can help define client needs. Our research focused on the question whether the facts obtained with Knowledge Discovery from Databases and Data Mining can provide us with useful information on client needs and whether the information can be used for sale campaigns in insurance companies. Results confirmed that besides already known facts also some new and surprising facts appear when the artificial intelligence methods are used. On the basis of these new facts we can prepare more appropriate individual offers for the client. The approach of individual offers results in more satisfied clients. Different methods based on artificial intelligence can lead to better efficiency and reduction of business costs