Identifying Semantically Deviating Outlier Documents

Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 2748–2757 Copenhagen, Denmark, September 7–11, 2017. c⃝2017 Association for Computational Linguistics Identifying Semantically Deviating Outlier Documents∗ Honglei Zhuang1, Chi Wang2, Fangbo Tao1, Lance Kaplan3 and Jiawei Han1 1Department of Computer Science, University of Illinois at Urbana-Champaign 2Microsoft Research, Redmond 3US Army Research Lab {hzhuang3, ftao2, hanj}@illinois.edu, wang.chi@microsoft.com, lance.m.kaplan.civ@mail.mil Abstract A document outlier is a document that sub- stantially deviates in semantics from the majority ones in a corpus. Automatic iden- tification of document outliers can be valu- able in many applications, such as screen- ing health records for medical mistakes. In this paper, we study the problem of mining semantically deviating document outliers in a given corpus. We develop a generative model to identify frequent and character- istic semantic regions in the word embed- ding space to represent the given corpus, and a robust outlierness measure which is resistant to noisy content in documents. Experiments conducted on two real-world textual data sets show that our method can achieve an up to 135% improvement over baselines in terms of recall at top-1% of the outlier ranking. 1 Introduction The technology today has made it unprecedent- edly easy to collect and store documents in an increasing number of domains. Automatic text analysis (e.g. document clustering, summariza- tion, topic modeling) becomes more useful and de- manded as the corpus size grows. Some trending ∗ Research was sponsored in part by the U.S. Army Re- search Lab. under Cooperative Agreement No. W911NF- 09-2-0053 (NSCTA), National Science Foundation IIS- 1320617, IIS 16-18481, and NSF IIS 17-04532, and grant 1U54GM114838 awarded by NIGMS through funds pro- vided by the trans-NIH Big Data to Knowledge (BD2K) initiative (www.bd2k.nih.gov). The views and conclusions contained in this document are those of the author(s) and should not be interpreted as representing the official policies of the U.S. Army Research Laboratory or the U.S. Govern- ment. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation hereon. domains (e.g. health records) call for a new analyt- ical task, mining outlier documents: given a cor- pus, identify a small number of documents which substantially deviate from the semantic focuses of the given corpus. Outlier documents can provide valuable insights or imply potential errors. For example, an outlier health record from records of the same disease could indicate a new variation of the disease if it has an abnormal symptom descrip- tion, or a medical error if it has an abnormal treat- ment description. A previous study (Hauskrecht et al., 2013) uses structured data in health records to show the importance of this application, and points out that further improvement should be achieved by leveraging text data. Existing work has studied a related albeit different task, novel document detection (Ka- siviswanathan et al., 2012, 2013; Zhang et al., 2002, 2004), where one aims to identify from a document stream if a newly arriving document is novel or redundant. In other words, this task …

Content truncated to 500 words