Hydra: An Open Framework for Virtual-Fusion of Recommendation Filters

Today’s web commercial applications demand more powerfulrecommendation systems due to the rapid increase in thenumber of both consumers and available products. Searchingfor the best algorithm with the highest accuracy and realisticcomplexity is, most of the time, a very costly processin terms of both time and resources. In this paper we suggestan alternative framework called Hydra which enablesthe virtual fusion of any and as many currently availablerecommendation algorithms in such a distributed mannerthat algorithms’ complexities are not summarized but parallelized.Therefore, we utilize the available algorithms andtechnologies aiming to achieve better accuracy in order tosurpass even the most state of the art algorithms. In addition,Hydra can be used to find how algorithms interactwith each other in order to estimate the resulting accuracytowards inventing a more precise algorithm diminishing therisk of a failed investment. Hydra can be adjusted and integratedin any recommendation application while it is alsoopen to new functionalities which can be embedded easilyand in a transparent manner.