QoS-based Trust Evaluation for Data Services as a Black Box

by   Senda Romdhani, et al.

This paper proposes a QoS-based trust evaluation model for black box data services. Under the black-box model, data services neither export (meta)-data about conditions in which they are deployed and collect and process data nor the quality of data they deliver. Therefore, the black-box model creates blind spots about the extent to which data providers can be trusted to be used to build target applications. The trust evaluation model for black box data services introduced in this paper originally combines QoS indicators, like service performance and data quality, to determine services trustworthiness. The paper also introduces DETECT: a Data sErvice as a black box Trust Evaluation arChitecTure, that validates our model. The trust model and its associated monitoring strategies have been assessed in experiments with representative case studies. The results demonstrate the feasibility and effectiveness of our solution.



page 1

page 2

page 3

page 4


Dice in the Black Box: User Experiences with an Inscrutable Algorithm

We demonstrate that users may be prone to place an inordinate amount of ...

Modeling and Control with Local Linearizing Nadaraya Watson Regression

Black box models of technical systems are purely descriptive. They do no...

Black-Box Assessment of Optical Spectrum Services

A spectral sweep process is introduced to discover performance issues in...

An Ethical Black Box for Social Robots: a draft Open Standard

This paper introduces a draft open standard for the robot equivalent of ...

Iterative Orthogonal Feature Projection for Diagnosing Bias in Black-Box Models

Predictive models are increasingly deployed for the purpose of determini...

Integrity Fingerprinting of DNN with Double Black-box Design and Verification

Cloud-enabled Machine Learning as a Service (MLaaS) has shown enormous p...

Rendition: Reclaiming what a black box takes away

The premise of our work is deceptively familiar: A black box f(·) has al...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.