loading
Papers Papers/2022 Papers Papers/2022

Research.Publish.Connect.

Paper

Paper Unlock

Authors: Ilenia Fronza ; Alberto Sillitti ; Giancarlo Succi and Jelena Vlasenko

Affiliation: Free University of Bolzano and Bozen, Italy

Keyword(s): Failure prediction, Survival analysis, Cox PH model, Event sequence data, Log files.

Related Ontology Subjects/Areas/Topics: Artificial Intelligence and Decision Support Systems ; Enterprise Information Systems ; Industrial Applications of Artificial Intelligence

Abstract: Developing software without failures is indeed important. Still, it is also important to detect as soon as possible when a running application is likely to fail, so that corrective actions can be taken. Following the guidelines of Agile Methods, the goal of our research is to develop a statistical prediction model for failures that does not require any additional effort on the side of the developers of an application; the key concept is that the developers concentrate on the code and we use the information that is naturally generated by the running application to assess whether an application is likely to fail. So the developers concentrate only on providing direct value to the customer and then the model takes care of informing the environment of the possible crash. The proposed model uses as input data that is commonly produced by developers: the log files. The statistical prediction model employed comes from biomedical studies about cancer survival prediction based on gene express ion profiles where gene expression measurements and survival times of previous patients are used to predict future patients' survival. One of the most prominent models is the Cox Proportional Hazards (PH) model. In this work, we draw a parallel between our context and the biomedical one; we consider types of operations as genes, and operations and their multiplicity in the sequence as expression profiles. Then, we identify signature operations applying the above mentioned Cox PH model. We perform a prototypical analysis using real-world data to assess the suitability of our approach. We estimate the confidence interval of our results using Bootstrap. (More)

CC BY-NC-ND 4.0

Sign In Guest: Register as new SciTePress user now for free.

Sign In SciTePress user: please login.

PDF ImageMy Papers

You are not signed in, therefore limits apply to your IP address 3.144.114.8

In the current month:
Recent papers: 100 available of 100 total
2+ years older papers: 200 available of 200 total

Paper citation in several formats:
Fronza, I.; Sillitti, A.; Succi, G. and Vlasenko, J. (2011). TOWARD A NON INVASIVE CONTROL OF APPLICATIONS - A Biomedical Approach to Failure Prediction. In Proceedings of the 13th International Conference on Enterprise Information Systems - Volume 1: ICEIS; ISBN 978-989-8425-54-6; ISSN 2184-4992, SciTePress, pages 83-91. DOI: 10.5220/0003493000830091

@conference{iceis11,
author={Ilenia Fronza. and Alberto Sillitti. and Giancarlo Succi. and Jelena Vlasenko.},
title={TOWARD A NON INVASIVE CONTROL OF APPLICATIONS - A Biomedical Approach to Failure Prediction},
booktitle={Proceedings of the 13th International Conference on Enterprise Information Systems - Volume 1: ICEIS},
year={2011},
pages={83-91},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0003493000830091},
isbn={978-989-8425-54-6},
issn={2184-4992},
}

TY - CONF

JO - Proceedings of the 13th International Conference on Enterprise Information Systems - Volume 1: ICEIS
TI - TOWARD A NON INVASIVE CONTROL OF APPLICATIONS - A Biomedical Approach to Failure Prediction
SN - 978-989-8425-54-6
IS - 2184-4992
AU - Fronza, I.
AU - Sillitti, A.
AU - Succi, G.
AU - Vlasenko, J.
PY - 2011
SP - 83
EP - 91
DO - 10.5220/0003493000830091
PB - SciTePress