专家寻找模型融合框架研究

ON FUSION FRAMEWORK FOR EXPERT FINDING MODELS

网络中存在着规模庞大的信息,搜索引擎如Google为网络海量信息的检索提供了有效的途径,但是许多潜藏的知识仍然无法被搜索到。而且,大量知识并未存储于文档或者数据库中,其中大部分仅存在于人脑中。对于网络中无法检索到的知识,则需要找到掌握这些知识的专家,并通过交流获取这些知识。目前专家寻找的方法有语言模型、主题模型等,这些方法各有优缺点。提出一种专家寻找模型融合框架,该框架可有效地将已有的专家寻找模型结合起来,从而提高专家寻找的精确度与鲁棒性。实验结果支持了这一结论。

In internet there is large-scale information and the search engines such as Google offer effective way to the retrieval of mass Websites information.However, many other kinds of latent knowledge can still not able to be searched.Furthermore, a great deal of knowledge does not store in documents or databases, most of them only indwell in the brain of human being.For those knowledge cannot be retrieved in Website, it needs to find the experts who master the knowledge so that we can communicate with them to acquire the knowledge. Current expert finding methods include language model, topic model and so on.Every method has its own pros and cons.In the paper we put forward a fusion framework for expert finding models, it can combine existing expert finding models together effectively and thereby improves the precision and robustness of expert finding.Experimental results also support the conclusion.