The huge amount of irrelevant and unimportant information have led to the need of using personalization in selecting the information which is relevant to searchers’ interest. Personalized faceted search has been a potential tool to support searchers to retrieve appropriate information effectively by navigating a list of selected multiple facets or categories based on the search results. To develop an effective personalized faceted search, the selection of relevant multiple facets is an important mechanism. Collaborative-based personalization was introduced for facet selection. Recently, Artificial Neural Network (ANN) has been reported that it performs better than other state-of-the-art Collaborative Filtering techniques for predicting single facet. However, analyzing the collaborative interests for multiple facets has not been studied. It is challenging if the interaction of the users on multiple facets is based on the information associated with the preferences of similar users over a group of multiple facets. This paper proposes an ANN-based facet predictive model that makes use of the collaborative-based personalization concept for multiple facet selection. The architecture of the proposed model is based on two suitable interaction schemes, the Early interaction and the Late interaction schemes. Based on experimental results, the performance was evaluated in terms of prediction accuracy and computation time. The results showed that the proposed model based on an effective interaction scheme obtained significant improvement on the prediction of personal interests on multiple facets.

大量无关紧要的信息导致需要使用个性化来选择与搜索者的兴趣相关的信息。个性化的分面搜索已成为一种潜在的工具，可通过基于搜索结果导航选定的多个分面或类别的列表来支持搜索者有效地检索适当的信息。为了发展有效的个性化多面搜索，选择相关的多个方面是重要的机制。引入了基于协作的个性化来进行方面选择。最近，据报道，人工神经网络（ANN）的性能优于其他先进的协作过滤技术，可预测单面。但是，尚未研究分析多个方面的协作兴趣。如果多个方面上的用户交互是基于与相似用户在一组多个方面上的偏好相关联的信息，则具有挑战性。本文提出了一种基于ANN的方面预测模型，该模型利用基于协作的个性化概念进行多方面选择。所提出的模型的体系结构基于两种合适的交互方案，即早期交互方案和晚期交互方案。根据实验结果，根据预测精度和计算时间评估了性能。结果表明，基于有效交互方案的模型在多方面预测个人兴趣方面取得了显着改进。