基于工作流程的推荐系统的设计

发布时间：2018-06-12 19:00

本文选题：推荐系统 + 工作流程　；参考：《重庆大学》2012年硕士论文

【摘要】：当今社会互联网技术已经得到广泛运用，这就带动了电子商务现今的高速发展，同时也导致了Internet中的资源数量以几何数级在快速增长。“信息爆炸”和“信息过载”使得人们在面对海量的信息内容时往往感到手足无措，他们很难从浩如烟海的信息中发现自己真正需要的资讯信息。人们提出了很多方法来试图解决信息过量问题，于是我们目前广泛使用的搜索引擎和信息检索技术应运而生。本文提出了一种基于工作流程的推荐系统模型，其目的是提供适当的信息给目标成员，主要针对团队协作的环境而不是日常生活（例如推荐商品、电影、新闻等）情况。相对于普通的日常生活情况，在团队合作的环境下可以更多的利用推荐系统获得信息。基于工作流程的协作团队包含有关成员之间的关系、角色和任务等信息，这就使得团队成员可以通过协同过滤获得所需要的信息。此外每个团队成员还可以通过工作流程安排表所提供的信息来获得自己所需要的适量的信息。本文中我们研究了基于工作流程的推荐系统运行的机制，并且进行了一系列团队协作实验，，这些实验验证了基于工作流程的推荐系统的有效性及其高效率。实验中我们发现，基于工作流程的协同过滤模式的推荐效果要优于普通的协同过滤模式，而考虑到任务之间关联的基于工作流程的协同过滤模式则是最优的。这些结果反映了协作环境可以提供成员之间相互关系的详细信息，以便提高协同过滤的效率。并且“顺序类型”的关系比“子类型”关系和“组成部分类型”关系对相应的成员具有更大的影响。我们还发现不同经验的团队成员会对推荐系统的推荐量产生影响。基于实验结果，我们认为我们所提出的基于工作流程的协同过滤方法，可能更适合于团队或组织成员之间存在相对较强的合作的情况。
[Abstract]:Nowadays, Internet technology has been widely used, which has led to the rapid development of electronic commerce, and also led to the rapid growth of the number of resources in the Internet. "Information explosion" and "information overload" make people often feel at a loss in the face of a great deal of information, and they find it difficult to find the information they really need from the vast amount of information. Many methods have been proposed to solve the problem of information overdose, so the search engine and information retrieval technology which are widely used at present come into being. In this paper, a workflow based recommendation system model is proposed, which aims to provide appropriate information to the target members, mainly in the context of teamwork rather than daily life (such as recommending products, movies, news, etc.). More information can be obtained from the recommendation system in a team-based environment than in the normal daily life. Workflow based collaborative teams contain information about relationships, roles and tasks among members, which enables team members to obtain the information they need through collaborative filtering. In addition, each team member can obtain the appropriate amount of information from the workflow schedule. In this paper, we study the mechanism of workflow based recommendation system, and carry out a series of team collaboration experiments, which verify the effectiveness and efficiency of workflow based recommendation system. In the experiment we find that the recommendation effect of the workflow based collaborative filtering model is better than that of the common collaborative filtering mode and the workflow based collaborative filtering model considering the correlation between tasks is the best. These results reflect that the collaborative environment can provide detailed information on the relationship between members in order to improve the efficiency of collaborative filtering. And the relation of "order type" has more influence on the corresponding member than the relation of "subtype" and "component type". We also found that team members with different experiences had an impact on the volume of referrals. Based on the experimental results, we think that the workflow based collaborative filtering method proposed by us may be more suitable for the situation where there is relatively strong cooperation between the team members and the members of the organization.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.3

【相似文献】