Contextual Outlier Interpretation

By Ninghao Liu, Donghwa Shin, Xia Hu. 2017

76 citations recorded by the source corpus.

While outlier detection has been intensively studied in many applications, interpretation is becoming increasingly important to help people trust and evaluate the developed detection models through providing intrinsic reasons why the given outliers are identified. It is a nontrivial task for interpreting the abnormality of outliers due to the distinct characteristics of different detection models, complicated structures of data in certain applications, and imbalanced distribution of outliers and normal instances. In addition, contexts where outliers locate, as well as the relation between outliers and the contexts, are usually overlooked in existing interpretation frameworks. To tackle the issues, in this paper, we propose a Contextual Outlier INterpretation (COIN) framework to explain the abnormality of outliers spotted by detectors. The interpretability of an outlier is achieved through three aspects, i.e., outlierness score, attributes that contribute to the abnormality, and contextual description of its neighborhoods. Experimental results on various types of datasets demonstrate the flexibility and effectiveness of the proposed framework.

Subfields: Artificial Intelligence, Computer Networks and Communications

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Xu Chen
2. Yongfeng Zhang
3. Xia Hu
4. Qingquan Song
5. Ninghao Liu
6. Mengnan Du
7. Lior Rokach
8. Bracha Shapira

In Computer Networks and Communications

1. Yongfeng Zhang
2. Xu Chen
3. Lior Rokach
4. Bracha Shapira
5. Xia Hu
6. Qingquan Song
7. Ninghao Liu

Citations in the corpus

• Explainable Recommendation: A Survey and New Perspectives (2018, methodology): “Lu et al. [2018] presented a deep learning recommendation model which co-learns user and item information from ratings and customer reviews by optimizing matrix factorization and an attention-based GRU network.”
• Explaining anomalies detected by autoencoders using Shapley Additive Explanations (2021, background): “Liu et al. [25] proposed a Contextual Outlier INterpretation (COIN) framework to explain anomalies spotted by detectors.”
• Towards Explanation of DNN-based Prediction with Guided Feature Inversion (2018, methodology): “There are a number of techniques developed to interpret machine learning models [7, 12, 23, 24, 31].”