ATD: Anomalous Topic Discovery in High Dimensional Discrete Data

Hossein Soleimani, David J. Miller

    Research output: Contribution to journalArticle

    4 Citations (Scopus)

    Abstract

    We propose an algorithm for detecting patterns exhibited by anomalous clusters in high dimensional discrete data. Unlike most anomaly detection (AD) methods, which detect individual anomalies, our proposed method detects groups ( clusters) of anomalies; i.e., sets of points which collectively exhibit abnormal patterns. In many applications, this can lead to a better understanding of the nature of the atypical behavior and to identifying the sources of the anomalies. Moreover, we consider the case where the atypical patterns exhibit on only a small (salient) subset of the very high dimensional feature space. Individual AD techniques and techniques that detect anomalies using all the features typically fail to detect such anomalies, but our method can detect such instances collectively, discover the shared anomalous patterns exhibited by them, and identify the subsets of salient features. In this paper, we focus on detecting anomalous topics in a batch of text documents, developing our algorithm based on topic models. Results of our experiments show that our method can accurately detect anomalous topics and salient features (words) under each such topic in a synthetic data set and two real-world text corpora and achieves better performance compared to both standard group AD and individual AD techniques. All required code to reproduce our experiments is available from https://github.com/hsoleimani/ATD.

    Original languageEnglish (US)
    Article number7463492
    Pages (from-to)2267-2280
    Number of pages14
    JournalIEEE Transactions on Knowledge and Data Engineering
    Volume28
    Issue number9
    DOIs
    StatePublished - Sep 1 2016

    Fingerprint

    Experiments

    All Science Journal Classification (ASJC) codes

    • Information Systems
    • Computer Science Applications
    • Computational Theory and Mathematics

    Cite this

    @article{7249d6bbef07420896592d5dfcfbf4df,
    title = "ATD: Anomalous Topic Discovery in High Dimensional Discrete Data",
    abstract = "We propose an algorithm for detecting patterns exhibited by anomalous clusters in high dimensional discrete data. Unlike most anomaly detection (AD) methods, which detect individual anomalies, our proposed method detects groups ( clusters) of anomalies; i.e., sets of points which collectively exhibit abnormal patterns. In many applications, this can lead to a better understanding of the nature of the atypical behavior and to identifying the sources of the anomalies. Moreover, we consider the case where the atypical patterns exhibit on only a small (salient) subset of the very high dimensional feature space. Individual AD techniques and techniques that detect anomalies using all the features typically fail to detect such anomalies, but our method can detect such instances collectively, discover the shared anomalous patterns exhibited by them, and identify the subsets of salient features. In this paper, we focus on detecting anomalous topics in a batch of text documents, developing our algorithm based on topic models. Results of our experiments show that our method can accurately detect anomalous topics and salient features (words) under each such topic in a synthetic data set and two real-world text corpora and achieves better performance compared to both standard group AD and individual AD techniques. All required code to reproduce our experiments is available from https://github.com/hsoleimani/ATD.",
    author = "Hossein Soleimani and Miller, {David J.}",
    year = "2016",
    month = "9",
    day = "1",
    doi = "10.1109/TKDE.2016.2561288",
    language = "English (US)",
    volume = "28",
    pages = "2267--2280",
    journal = "IEEE Transactions on Knowledge and Data Engineering",
    issn = "1041-4347",
    publisher = "IEEE Computer Society",
    number = "9",

    }

    ATD : Anomalous Topic Discovery in High Dimensional Discrete Data. / Soleimani, Hossein; Miller, David J.

    In: IEEE Transactions on Knowledge and Data Engineering, Vol. 28, No. 9, 7463492, 01.09.2016, p. 2267-2280.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - ATD

    T2 - Anomalous Topic Discovery in High Dimensional Discrete Data

    AU - Soleimani, Hossein

    AU - Miller, David J.

    PY - 2016/9/1

    Y1 - 2016/9/1

    N2 - We propose an algorithm for detecting patterns exhibited by anomalous clusters in high dimensional discrete data. Unlike most anomaly detection (AD) methods, which detect individual anomalies, our proposed method detects groups ( clusters) of anomalies; i.e., sets of points which collectively exhibit abnormal patterns. In many applications, this can lead to a better understanding of the nature of the atypical behavior and to identifying the sources of the anomalies. Moreover, we consider the case where the atypical patterns exhibit on only a small (salient) subset of the very high dimensional feature space. Individual AD techniques and techniques that detect anomalies using all the features typically fail to detect such anomalies, but our method can detect such instances collectively, discover the shared anomalous patterns exhibited by them, and identify the subsets of salient features. In this paper, we focus on detecting anomalous topics in a batch of text documents, developing our algorithm based on topic models. Results of our experiments show that our method can accurately detect anomalous topics and salient features (words) under each such topic in a synthetic data set and two real-world text corpora and achieves better performance compared to both standard group AD and individual AD techniques. All required code to reproduce our experiments is available from https://github.com/hsoleimani/ATD.

    AB - We propose an algorithm for detecting patterns exhibited by anomalous clusters in high dimensional discrete data. Unlike most anomaly detection (AD) methods, which detect individual anomalies, our proposed method detects groups ( clusters) of anomalies; i.e., sets of points which collectively exhibit abnormal patterns. In many applications, this can lead to a better understanding of the nature of the atypical behavior and to identifying the sources of the anomalies. Moreover, we consider the case where the atypical patterns exhibit on only a small (salient) subset of the very high dimensional feature space. Individual AD techniques and techniques that detect anomalies using all the features typically fail to detect such anomalies, but our method can detect such instances collectively, discover the shared anomalous patterns exhibited by them, and identify the subsets of salient features. In this paper, we focus on detecting anomalous topics in a batch of text documents, developing our algorithm based on topic models. Results of our experiments show that our method can accurately detect anomalous topics and salient features (words) under each such topic in a synthetic data set and two real-world text corpora and achieves better performance compared to both standard group AD and individual AD techniques. All required code to reproduce our experiments is available from https://github.com/hsoleimani/ATD.

    UR - http://www.scopus.com/inward/record.url?scp=84982180045&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84982180045&partnerID=8YFLogxK

    U2 - 10.1109/TKDE.2016.2561288

    DO - 10.1109/TKDE.2016.2561288

    M3 - Article

    AN - SCOPUS:84982180045

    VL - 28

    SP - 2267

    EP - 2280

    JO - IEEE Transactions on Knowledge and Data Engineering

    JF - IEEE Transactions on Knowledge and Data Engineering

    SN - 1041-4347

    IS - 9

    M1 - 7463492

    ER -