PROJECT TITLE :

HAM Hybrid Associations Models for Sequential Recommendation

ABSTRACT:

The goal of sequential recommendation is to determine and suggest to a user the next few items that the user is most likely to purchase or rate, taking into account the user's historical purchasing and rating patterns. It transforms into an efficient instrument that assists users in selecting preferred items from among a number of available options. In this manuscript, we developed hybrid associations models (HAM) to generate sequential recommendations based on the following three factors: 1) the users' long-term preferences; 2) sequential, high-order and low-order association patterns in the users' most recent purchases/ratings; and 3) synergies among those items. HAM uses element-wise product to represent item synergies of arbitrary orders, and simplistic pooling to represent a set of items in the associations. Both of these methods are described in more detail below. On a total of six public benchmark datasets and in three distinct experimental conditions, we compared HAM models to the most recent methods that are considered to be state-of-the-art. The results of our experiments show that HAM models perform significantly better than the current state of the art in all of the experimental settings we tested them in. with an increase of up to 46.6 percent in overall quality. In addition, the results of our comparison of run-time performance during testing show that HAM models are significantly more effective than the methods that are considered to be state-of-the-art. and are able to accomplish significant speedups of up to 139.7 folds in some cases.