中文摘要：

由于群体成员个性不同，尤其在群成员意见冲突情况下如何准确获取群体成员偏好是群体推荐系统进行有效推荐的关键。针对上述问题，充分考虑群体成员在冲突情境下的行为特征，采用Thomas-Kilmann行为冲突模式分析方法（TKI）衡量群成员在冲突情境下对不同偏好的接受度；并将群体成员接受度和成员一群体相似度有机融合到传统群体推荐算法中，提出了一种新的混合群体推荐算法。实验结果表明，新算法比传统群体推荐算法的推荐精度有较为明显的改善。

英文摘要：

As personality varies from person to person, it was crucial for group recommendation to obtain the exact preference of group members especially when conflict of opinions existed among them. To response to this situation, this paper took into account the personality of group members and analyzed their acceptance degree to conflicting opinions by using the Thomas-Kilmann conflict mode instrument（TKI）. Then it proposed a novel hybrid algorithm by considering the acceptance degree of group members along with their similarity to the group. An experiment with 185 participants is conducted. The results show that the proposed algorithm has significant improvement compared to benchmark methods.