@article{fdi:010086137,
  title = {{S}intel : a machine learning framework to extract insights from signals},
  author = {{A}lnegheimish, {S}. and {L}iu, {D}. {Y}. and {S}ala, {C}. and {B}erti-{E}quille, {L}aure and {V}eeramachaneni, {K}. and {A}cm,},
  editor = {},
  language = {{ENG}},
  abstract = {{T}he detection of anomalies in time series data is a critical task with many monitoring applications. {E}xisting systems often fail to encompass an end-to-end detection process, to facilitate comparative analysis of various anomaly detection methods, or to incorporate human knowledge to refine output. {T}his precludes current methods from being used in real-world settings by practitioners who are not {ML} experts. {I}n this paper, we introduce {S}intel, a machine learning framework for end-to-end time series tasks such as anomaly detection. {T}he framework uses state-of-the-art approaches to support all steps of the anomaly detection process. {S}intel logs the entire anomaly detection journey, providing detailed documentation of anomalies over time. {I}t enables users to analyze signals, compare methods, and investigate anomalies through an interactive visualization tool, where they can annotate, modify, create, and remove events. {U}sing these annotations, the framework leverages human knowledge to improve the anomaly detection pipeline. {W}e demonstrate the usability, efficiency, and effectiveness of {S}intel through a series of experiments on three public time series datasets, as well as one real-world use case involving spacecraft experts tasked with anomaly analysis tasks. {S}inters framework, code, and datasets are open-sourced at https://github.com/sintel-dev/.},
  keywords = {{M}achine {L}earning {F}ramework ; {A}nomaly {D}etection ; {H}uman-{I}n-the-{L}oop {AI} ; {T}ime {S}eries {D}ata ; {D}ata {S}cience {P}ipeline},
  booktitle = {},
  journal = {{P}roceedings of the 2022 {I}nternational {C}onference on {M}anagement of {D}ata},
  numero = {},
  pages = {1855--1865},
  ISSN = {0730-8078},
  year = {2022},
  DOI = {10.1145/3514221.3517910},
  URL = {https://www.documentation.ird.fr/hor/fdi:010086137},
}