FRIOD: a deeply integrated feature-rich interactive system for effective and efficient outlier detection

Zhu, Xiaodong and Zhang, Ji and Li, Hongzhou and Fournier-Viger, Philippe and Lin, Jerry Chun-Wei and Chang, Liang (2017) FRIOD: a deeply integrated feature-rich interactive system for effective and efficient outlier detection. IEEE Access, 5:8101452. pp. 25682-25695.

[img]
Preview
Text (Accepted Version)
FINAL Article.pdf

Download (1MB) | Preview

Abstract

In this paper, we propose an novel interactive outlier detection system called feature-rich interactive outlier detection (FRIOD), which features a deep integration of human interaction to improve detection performance and greatly streamline the detection process. A user-friendly interactive mechanism is developed to allow easy and intuitive user interaction in all the major stages of the underlying outlier detection algorithm which includes dense cell selection, location-aware distance thresholding, and final top outlier validation. By doing so, we can mitigate the major difficulty of the competitive outlier detection methods in specifying the key parameter values, such as the density and distance thresholds. An innovative optimization approach is also proposed to optimize the grid-based space partitioning, which is a critical step of FRIOD. Such optimization fully considers the high-quality outliers it detects with the aid of human interaction. The experimental evaluation demonstrates that FRIOD can improve the quality of the detected outliers and make the detection process more intuitive, effective, and efficient.


Statistics for USQ ePrint 36145
Statistics for this ePrint Item
Item Type: Article (Commonwealth Reporting Category C)
Refereed: Yes
Item Status: Live Archive
Faculty/School / Institute/Centre: Historic - Faculty of Health, Engineering and Sciences - School of Agricultural, Computational and Environmental Sciences (1 Jul 2013 - 5 Sep 2019)
Faculty/School / Institute/Centre: Historic - Faculty of Health, Engineering and Sciences - School of Agricultural, Computational and Environmental Sciences (1 Jul 2013 - 5 Sep 2019)
Date Deposited: 20 Mar 2020 06:05
Last Modified: 08 Jun 2021 03:02
Uncontrolled Keywords: human interaction, outlier detection, space partitioning visualization
Fields of Research (2008): 08 Information and Computing Sciences > 0801 Artificial Intelligence and Image Processing > 080109 Pattern Recognition and Data Mining
Identification Number or DOI: https://doi.org/10.1109/ACCESS.2017.2771237
URI: http://eprints.usq.edu.au/id/eprint/36145

Actions (login required)

View Item Archive Repository Staff Only