Explainable Artificial Intelligence for Process Mining: A General Overview and Application of a Novel Local Explanation Approach for Predictive Process Monitoring

Mehdiyev, Nijat; Fettke, Peter

doi:10.1007/978-3-030-64949-4_1

Cited by 40 publications

(20 citation statements)

References 86 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A wide range of works have already evaluated these different models with performance-based metrics [3,8,26]. Over the last two years, there has also been a lot of movement around explainability in predictive process monitoring [23,27,22,28,29,30,11,19]. Related to the XAI literature [13], these different works can be divided into two different trends based on how they deal with the explainability-accuracy trade-off.…”

Section: Related Work and Motivationmentioning

confidence: 99%

“…The first trend presumes the complex model as the task model [14] and looks for explanations using post-hoc techniques. In predictive process monitoring, several papers have already suggested model-agnostic explainability techniques on top of machine learning models [23,22] such as SHapley Additive exPlanations (SHAP) [31] or Local Interpretable Model-Agnostic Explanations (LIME) [32], with similar developments in a deep learning context [27,30,33,28,29]. In [27],…”

Section: Related Work and Motivationmentioning

confidence: 99%

“…an LSTM model in combination with SHAP values allow the user to identify the influence of certain attributes in the different steps of the process, while Mehdiyev focuses on creating local post-hoc explanations with the use of a surrogate decision tree [30]. In [33] and [28], they use a Bidirectional LSTM (one forward RNN and one backward RNN), where after the hidden states of the time steps of both RNNs are concatenated.…”

Section: Related Work and Motivationmentioning

confidence: 99%

“…Explainability model. An explainability model is the umbrella term for the post-hoc explainability techniques that exist to generate explanations for a black box model such as surrogate models [30], attention layers [28] and SHAP values [43]. Previous research has mostly referred to this as post-hoc explanations [14,12], or post-hoc interpretations [39].…”

Section: The Different Xai Nomenclaturesmentioning

confidence: 99%

See 3 more Smart Citations

Explainable Predictive Process Monitoring: Evaluation Metrics and Guidelines for Process Outcome Prediction

Stevens¹,

Smedt²

2022

Preprint

View full text Add to dashboard Cite

Recently, a shift has been made in the field of Outcome-Oriented Predictive Process Monitoring (OOPPM) to use models from the eXplainable Artificial Intelligence paradigm, however the evaluation still occurs mainly through performancebased metrics not accounting for the implications and lack of actionability of the explanations. In this paper, we define explainability by the interpretability of the explanations (through the widely-used XAI properties parsimony and functional complexity) and the faithfulness of the explainability model (through monotonicity and level of disagreement). The introduced properties are analysed along the event, case, and control flow perspective that are typical of a process-based analysis. This allows to quantitatively compare, inter alia, inherently created explanations (e.g., logistic regression coefficients) with post-hoc explanations (e.g., Shapley values). Moreover, this paper contributes a guideline named X-MOP to practitioners to select the appropriate model based on the event log specifications and the task at hand, by providing insight into how the varying preprocessing, model complexity and post-hoc explainability techniques typical in OOPPM influence the explainability of the model. To this end, we benchmark seven classifiers on thirteen real-life events logs.

show abstract

Section: Related Work and Motivationmentioning

confidence: 99%

Section: Related Work and Motivationmentioning

confidence: 99%

Section: Related Work and Motivationmentioning

confidence: 99%

Section: The Different Xai Nomenclaturesmentioning

confidence: 99%

See 2 more Smart Citations

Explainable Predictive Process Monitoring: Evaluation Metrics and Guidelines for Process Outcome Prediction

Stevens¹,

Smedt²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Most of the existing work in Explainable deep learning-based predictive process analytics approaches use post-hoc methods such as LIME and SHAP to explain the model's prediction [33,22], while some of the recent approaches focus on intrinsic interpretable deep learning architectures. These include the approaches that use model attention [33], explicit process model and Gated graph neural network based approaches [34], and use of partial dependence plots [35] and Layer wise relevance propagation [15].…”

Section: Deep Learning-based Predictive Process Analyticsmentioning

confidence: 99%

Building Interpretable Models for Business Process Prediction using Shared and Specialised Attention Mechanisms

Wickramanayake,

He,

Ouyang

et al. 2021

Preprint

View full text Add to dashboard Cite

Predictive process analytics, often underpinned by deep learning techniques, is a newly emerged discipline dedicated to providing business process intelligence in modern organisations. Whilst accuracy has been a dominant criterion in building predictive capabilities, the use of deep learning techniques comes at the cost of the resulting models being used as 'black boxes', i.e., they are unable to provide insights into why a certain business process prediction was made. So far little attention has been paid to supporting interpretability in the design of deep learning-based process predictive models.In this paper, we address the 'black-box' problem in the context of predictive process analytics by developing attention-based models that are capable to inform both what and why is a prediction. We propose i) two types of attentions -event attention to capture the impact of specific events on a prediction, and attribute attention to reveal which attribute(s) of an event influenced the prediction; and ii) two attention mechanisms -shared attention mechanism and specialised attention mechanism to reflect different design decisions between whether to construct attribute attention on individual input features (specialised) or using the concatenated feature tensor of all input feature vectors (shared). These lead to two distinct attentionbased models, and both are interpretable models that incorporate interpretability directly into the structure of a process predictive model. We conduct experimental evaluation of the proposed models using real-life dataset, and comparative analysis between the models for accuracy and interpretability.The results demonstrate that i) the proposed attention-based models can achieve reasonably high accuracy; ii) both are capable of providing relevant interpretations (when validated against domain knowledge); and iii) while the shared attention-based model tends to perform (slightly) better than the specialised attention-based model, the latter tends to provide more relevant interpretations than the former, reflecting the fact that the specialised attention-based model is designed to facilitate better interpretability.

show abstract

Explainable Artificial Intelligence (XAI) Supporting Public Administration Processes – On the Potential of XAI in Tax Audit Processes

Mehdiyev

Houy

Gutermuth

et al. 2021

Lecture Notes in Information Systems and Organisation

View full text Add to dashboard Cite

Explainable Artificial Intelligence for Process Mining: A General Overview and Application of a Novel Local Explanation Approach for Predictive Process Monitoring

Abstract: to the edited volume "Interpretable Artificial Intelligence: A perspective of Granular Computing" (published by Springer).

Cited by 40 publications

References 86 publications

Explainable Predictive Process Monitoring: Evaluation Metrics and Guidelines for Process Outcome Prediction

Explainable Predictive Process Monitoring: Evaluation Metrics and Guidelines for Process Outcome Prediction

Building Interpretable Models for Business Process Prediction using Shared and Specialised Attention Mechanisms

Explainable Artificial Intelligence (XAI) Supporting Public Administration Processes – On the Potential of XAI in Tax Audit Processes

Contact Info

Product

Resources

About