Factual and Counterfactual Explanations for Black Box Decision Making
Open accessThe DOI matches its Crossref record; 12 of the 14 references it cites were found in a registry or an index.
Abstract
The rise of sophisticated machine learning models has brought accurate but obscure decision systems, which hide their logic, thus undermining transparency, trust, and the adoption of artificial intelligence (AI) in socially sensitive and safety-critical contexts. We introduce a local rule-based explanation method, providing faithful explanations of the decision made by a black box classifier on a specific instance. The proposed method first learns an interpretable, local classifier on a synthetic neighborhood of the instance under investigation, generated by a genetic algorithm. Then, it derives from the interpretable classifier an explanation consisting of a decision rule, explaining the factual reasons of the decision, and a set of counterfactuals, suggesting the changes in the instance features that would lead to a different outcome. Experimental results show that the proposed method outperforms existing approaches in terms of the quality of the explanations and of the accuracy in mimicking the black box.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
"Why Should I Trust You?": Explaining the Predictions of Any Classifier
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2939778 · Cited by 17,555
Explanation in artificial intelligence: Insights from the social sciences
Miller, T. (2019) · Artificial Intelligence · doi:10.1016/j.artint.2018.07.007 · Cited by 5,302
Anchors: High-Precision Model-Agnostic Explanations
Ribeiro, M. T., Singh, S., & Guestrin, C. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.11491 · Cited by 2,162 · cited 1×
“We also evaluate the rule coverage, namely the fraction of instances to explain covered by the rule: large values of coverage means better rule generalization.(4) Figure 8 also shows the average coverage of the decision rules, with LORE showing a consistently better coverage than ANCHOR.”
Counterfactual explanations without opening the black box: automated decisions and the GDPR
Wachter, S., Mittelstadt, B., & Russell, C. (2017) · SSRN Electronic Journal · doi:10.2139/ssrn.3063289 · Cited by 1,061
Evolutionary Computation 1: Basic Algorithms and Operators
(2018) · CRC Press eBooks · doi:10.1201/9781482268713 · Cited by 634
A Survey of Methods for Explaining Black Box Models
Guidotti, R., Monreale, A., Ruggieri, S., Turini, F., Giannotti, F., & Pedreschi, D. (2019) · ACM Computing Surveys · doi:10.1145/3236009 · Cited by 629
Self-citationEvolutionary Computation 2: Advanced Algorithms and Operators
(2000) · Taylor & Francis eBooks · doi:10.1201/9781420034349 · Cited by 451
A review of instance selection methods
Olvera-López, J. A., Carrasco-Ochoa, J. A., Martínez-Trinidad, J. F., & Kittler, J. (2010) · Artificial Intelligence Review · doi:10.1007/s10462-010-9165-y · Cited by 394 · cited 1×
“This task differs from approaches to instance selection based on genetic algorithms.(7) In our case, we cannot assume the availability of the training set of b, or not even that b is a supervised ML predictor for which a training set exists.”
MethodCounterfactuals in Explainable Artificial Intelligence (XAI): Evidence from Human Reasoning
Byrne, R. M. J. (2019) · Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/876 · Cited by 285 · cited 1×
“the cognitive psychology literature maintaining that counterfactuals help people to reason on explanations that identify cause-effect or reason-action relations between events.(5,6) Similarly, Grover et al.”
BackgroundA Survey on Evolutionary Instance Selection and Generation
Derrac, J., García, S., & Herrera, F. (2010) · International Journal of Applied Metaheuristic Computing · doi:10.4018/jamc.2010102604 · Cited by 75 · cited 1×
“Our task is instead similar to instance generation in the field of active learning, including evolutionary approaches.(8) We adopt an approach based on a genetic algorithm, which generates z 2 Z1⁄4 [ Z61⁄4 by maximizing the following fitness functions:”
Method
Cited by
Items 201–300 of 438
Szczuka, J. M., Horstmann, A. C., Szymczyk, N., Strathmann, C., Artelt, A., Mavrina, L., & Krämer, N. C. (2024) · Nordic Conference on Human-Computer Interaction · doi:10.1145/3679318.3685351 · Cited by 8
“Contrastive explanations (also known as counterfactual explanations [2, 15, 54]), on the other hand, explain what changes in the input would have led to a different classification.”
Fundamentals on explainable and interpretable artificial intelligence models
Ducange, P., Marcelloni, F., Renda, A., & Ruffini, F. (2025) · Elsevier eBooks · doi:10.1016/b978-0-44-323761-4.00025-0 · Cited by 8
Colmenarejo, A. B., State, L., & Comandé, G. (2025) · International Review of Law Computers & Technology · doi:10.1080/13600869.2025.2497633 · Cited by 8
Influence of context on users’ views about explanations for decision-tree predictions
Maruf, S., Zukerman, I., Reiter, E., & Haffari, G. (2023) · Computer Speech & Language · doi:10.1016/j.csl.2023.101483 · Cited by 8
Maaroof, N., Moreno, A., Valls, A., Jabreel, M., & Romero‐Aroca, P. (2023) · Electronics · doi:10.3390/electronics12102215 · Cited by 8
Data-Agnostic Local Neighborhood Generation
Guidotti, R., & Monreale, A. (2020) · Proceedings · doi:10.1109/icdm50108.2020.00122 · Cited by 8
“Different works of the literature [5], [6], [8] highlighted the advantages of using “local” synthetic data in the learning phase of a predictive model.”
BackgroundMethodSelf-citationZhang, X., Wang, Y., & Li, Z. (2021) · Applied Intelligence · doi:10.1007/s10489-021-02255-z · Cited by 8
Yulianti, H., & Pranoto, G. T. (2021) · JISA(Jurnal Informatika dan Sains) · doi:10.31326/jisa.v4i2.1045 · Cited by 8
“This test ensures whether the process carried out produces output that is appropriate or not in accordance with the design.[17]”
ResultLearning Dashboards for Academic Advising in Practice
Laet, T. D. (2023) · Advances in analytics for learning and teaching · doi:10.1007/978-3-031-27646-0_4 · Cited by 8
Design for Interpretability: Meeting the Certification Challenge for Surgical Robots
Fiazza, M.-C., & Fiorini, P. (2021) · IEEE International Conference on Intelligence and Safety for Robotics (ISR) · doi:10.1109/isr50024.2021.9419378 · Cited by 8
“in [10]) 2) encapsulating a white box and restricting ANNs to using only interpretable primitives to approximate general functions (model-based interpretability, e.”
MethodFauvel, K., Masson, V., & Fromont, É. (2020) · arXiv (Cornell University) · Cited by 8
iXGB: Improving the Interpretability of XGBoost Using Decision Rules and Counterfactuals
Islam, M. R., Ahmed, M. U., & Begum, S. (2024) · International Conference on Agents and Artificial Intelligence (ICAART) · doi:10.5220/0012474000003636 · Cited by 7
“The coverage values of rules for classification models are expected to be higher for better generalisation (Guidotti et al., 2019).”
BackgroundNavigating the Structured What-If Spaces: Counterfactual Generation via Structured Diffusion
Madaan, N., & Bedathur, S. (2024) · IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) · doi:10.1109/satml59370.2024.00041 · Cited by 7
“For the tabular domain, various studies have pursued counterfactual explanations [6], [7], [9], [33], [34].”
BackgroundExploring Large Language Models Capabilities to Explain Decision Trees
Serafim, P. B. S., Crescenzi, P., Gezici, G., Cappuccio, E., Rinzivillo, S., & Giannotti, F. (2024) · Frontiers in artificial intelligence and applications · doi:10.3233/faia240183 · Cited by 7
Self-citationAn investigation into creating counterfactual examples for non-linear Support Vector Machines
Bergamin, L., & Aiolli, F. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.130809 · Cited by 7
Local interpretation of supervised learning models based on high dimensional model representation
Zhang, X., Ling, W., & Li, Z. (2021) · Information Sciences · doi:10.1016/j.ins.2021.01.079 · Cited by 7
EXPHLOT: EXplainable Privacy Assessment for Human LOcation Trajectories
Naretto, F., Pellungrini, R., Rinzivillo, S., & Fadda, D. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-45275-8_22 · Cited by 7
Evaluating Transparency: A Cross-Model Exploration of Explainable AI in Financial Forecasting
Valensky, D., & Mohaghegh, M. (2023) · IEEE Asia-Pacific Conference on Computer Science and Data Engineering (CSDE) · doi:10.1109/csde59766.2023.10487732 · Cited by 8
“The most common model-agnostic methods are LIME [24], SHAP [18], Anchors [23] and LoRE [12].”
MethodEngineering a Digital Twin for Diagnosis and Treatment of Multiple Sclerosis
d’Aloisio, G., Matteo, A. D., Cipriani, A., Lozzi, D., Mattei, E., Zanfardino, G., Marco, A. D., & Placidi, G. (2024) · ACM/IEEE International Conference on Model Driven Engineering Languages and Systems (MODELS) · doi:10.1145/3652620.3688249 · Cited by 7
“Hence, as also highlighted by the recently approved AI Act from the European Union [17], special attention must be devoted to ensuring the ethical quality of those AI systems (i.e., their fairness [11, 15, 16], privacy [13], and explainability [22, 23]).”
Communicating Uncertainty in Explanations of the Outcomes of Machine Learning Models
Zukerman, I., & Maruf, S. (2024) · International Conference on Natural Language Generation · doi:10.18653/v1/2024.inlg-main.4 · Cited by 7
“The explanations we showed in this study were based on the feature values in the path between the root of the decision tree and a prediction (Guidotti et al., 2019; Stepin et al., 2020).”
MethodYulianti, H. (2020) · MULTINETICS · doi:10.32722/multinetics.v6i2.3441 · Cited by 7
An Open-Source Software Library for Explainable Support Vector Machine Classification
Loor, M., Tapia-Rosero, A., & Tré, G. D. (2022) · IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) · doi:10.1109/fuzz-ieee55066.2022.9882731 · Cited by 7
Exemplars and Counterexemplars Explanations for Skin Lesion Classifiers
Metta, C., Guidotti, R., Yin, Y., Gallinari, P., & Rinzivillo, S. (2022) · Frontiers in artificial intelligence and applications · doi:10.3233/faia220209 · Cited by 7
“First, ABELE generates a neighborhood exploiting an AAE [12], then it learns a decision tree on the latent neighborhood providing local decision and counterfactual rules [14], and finally selects and decodes exemplars and counter-exemplars, extracting a saliency map.”
MethodSelf-citationKırboğa, K. K., Küçüksille, E. U., & Köse, U. (2023) · BRAIN BROAD RESEARCH IN ARTIFICIAL INTELLIGENCE AND NEUROSCIENCE · doi:10.18662/brain/14.3/475 · Cited by 7
“Several features are advantageous in our SHAP selection among many explainability techniques Guidotti et al. (2019);”
BackgroundTREX: Tree-Ensemble Representer-Point ExplanationsPreprint
Brophy, J., & Lowd, D. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2009.05530 · Cited by 7
“…perspective (Kazemitabar et al. 2017) as well as a local perspective on specific model predictions; these typically pertain to feature-attribution methods (Saabas 2014; Ribeiro, Singh, and Guestrin 2016; Lundberg and Lee 2017; Lundberg, Erion, and Lee 2018) or counterfactuals (Guidotti et al. 2019).”
BackgroundLocal Universal Rule-based eXplainer (LUX)
Bobek, S., & Nalepa, G. J. (2025) · SoftwareX · doi:10.1016/j.softx.2025.102102 · Cited by 7
Explaining Expert Search and Team Formation Systems with ExES
Golzadeh, K., Golab, L., & Szlichta, J. (2025) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde65448.2025.00024 · Cited by 6
LLiMe: enhancing text classifier explanations with large language models
Angiulli, F., Luca, F. D., Fassetti, F., & Nisticò, S. (2025) · Machine Learning · doi:10.1007/s10994-025-06920-z · Cited by 6
“…of each feature but is instead an example (or more) similar to the one considered by the explanation that, according to the black-box model, belongs to another class (Karimi et al., 2020; Mothilal et al., 2020; Laugel et al., 2018; Guidotti et al., 2019; Goyal et al., 2019; Angiulli et al., 2021a).”
Unified counterfactual explainer for graph neural networks
Giorgi, F., Silvestri, F., & Tolomei, G. (2026) · npj Artificial Intelligence · doi:10.1038/s44387-026-00135-w · Cited by 6
“Validity ( Ω ): also known as correctness, as de fi ned by Guidotti et al. 20,21 , indicates whether the explainer is capable of producing a valid counterfactual explanation.”
Contextual Boosting to Explainable SVM Classification
Loor, M., Tapia-Rosero, A., & Tré, G. D. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-39965-7_40 · Cited by 6
Counterfactual and Prototypical Explanations for Tabular Data via Interpretable Latent Space
Piaggesi, S., Bodria, F., Guidotti, R., Giannotti, F., & Pedreschi, D. (2024) · IEEE Access · doi:10.1109/access.2024.3496114 · Cited by 6
“For LORE [44], we exploit the genetic algorithm to generate 300 neighborhood samples.”
InfluentialSelf-citationPrediction and Explanation of Privacy Risk on Mobility Data with Neural Networks
Naretto, F., Pellungrini, R., Nardini, F. M., & Giannotti, F. (2020) · Communications in computer and information science · doi:10.1007/978-3-030-65965-3_34 · Cited by 6
Self-citationDesigning Shapelets for Interpretable Data-Agnostic Classification
Guidotti, R., & Monreale, A. (2021) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1145/3461702.3462553 · Cited by 6
“The lore explanation method [15] instead exploits a decision tree to return a”
BackgroundMethodSelf-citationShang, J., Xiao, Z., Tao, T., Wang, J., & Wu, Z. (2024) · Journal of King Saud University - Computer and Information Sciences · doi:10.1016/j.jksuci.2024.101962 · Cited by 6
“The improved method (Guidotti et al., 2019) uses genetic algorithm to accelerate its complex optimization process.”
Rischi etico-legali dell’Intelligenza Artificiale
Monreale, A. (2020) · DPCE Online · Cited by 6
Self-citationA System's Approach Taxonomy for User-Centred XAI: A SurveyPreprint
Emamirad, E., Omran, P. G., Haller, A., & Gregor, S. D. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2303.02810 · Cited by 6
“There are growing number of approaches aimed at effectively realising XAI, see e.g., [Guidotti et al. , 2019b; Gilpin et al. , 2018; Miller, 2019; Adadi and Berrada, 2018; Vilone and Longo, 2020; Zhang et al. , 2022].”
BackgroundBobek, S., & Nalepa, G. J. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2310.14894 · Cited by 6
“…most representative examples of the former are SHAP [19] or LIME [20], while the latter are mostly dominated by well established statistical methods such as partial dependence plots, accumulated local effects, functional decomposition, importance of permutation features or prototype discovery [21].”
MethodCIRF: Importance of related features for plausible counterfactual explanations
Kim, H.-D., Ju, Y.-J., Hong, J.-H., & Lee, S. (2024) · Information Sciences · doi:10.1016/j.ins.2024.120974 · Cited by 5
MASCOTS: Model-Agnostic Symbolic COunterfactual Explanations for Time Series
Płudowski, D., Spinnato, F., Wilczyński, P., Kotowski, K., Ntagiou, E., Guidotti, R., & Biecek, P. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-06078-5_6 · Cited by 5
Self-citationDiCoFlex: Model-Agnostic Diverse Counterfactuals with Flexible Control
Furman, O., Movsum-zada, U., Marszałek, P., Zięba, M., & Śmieja, M. (2025) · neural information processing systems · doi:10.52202/085713-2405 · Cited by 5
LORE4GroupRS: Explaining Group Recommendations Supported by a Local Rule-Based Approach
Yera, R., & Martínez, L. M. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-77738-7_25 · Cited by 5
Use of Artificial Intelligence and Machine Learning in Critical Care Ultrasound
Peck, M., & Conway, H. (2025) · Critical Care Clinics · doi:10.1016/j.ccc.2025.02.008 · Cited by 5
Xia, Z., & Gong, F. (2025) · Science China Mathematics · doi:10.1007/s11425-024-2454-9 · Cited by 5
“Chang et al. [7], Guidotti et al. [21,22] and Yu et al. [59] observed how the predicted results of the network convert among different synthetic examples to explain the results.”
Advancing Edge Artificial Intelligence: System Contexts
Vermesan, O., & Marples, D. (2024) · River Publishers eBooks · doi:10.1201/9781003478713 · Cited by 5
Relevant Irrelevance: Generating Alterfactual Explanations for Image Classifiers
Mertes, S., Huber, T., Karle, C., Weitz, K., Schlagowski, R., Conati, C., & André, E. (2024) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2024/52 · Cited by 5
Interpretable Latent Space to Enable Counterfactual Explanations
Bodria, F., Guidotti, R., Giannotti, F., & Pedreschi, D. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-18840-4_37 · Cited by 5
Self-citationDual-Track Lifelong Machine Learning-Based Fine-Grained Product Quality Analysis
Hong, X., Guan, S.-U., Xue, N., Li, Z., Man, K. L., Wong, P. W. H., & Liu, D. (2023) · Applied Sciences · doi:10.3390/app13031241 · Cited by 5
BELLA: Black box model Explanations by Local Linear Approximations
Radulovic, N., Bifet, A., & Suchanek, F. M. (2023) · Transactions on machine learning research · doi:10.48550/arxiv.2305.11311 · Cited by 5
“In addition to these standard metrics, one line of research argues that good explanations should be contrastive/counterfactual (Miller, 2018; Wachter, Mittelstadt, & Russell, 2017).”
Counterfactual Metarules for Local and Global RecoursePreprint
Bewley, T., Amoukou, S. I., Mishra, S., Magazzeni, D., & Veloso, M. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2405.18875 · Cited by 5
“We are aware of one prior method that uses rules to provide diverse CEs for groups of instances comparably to our rule/metarule formulation (Rawal & Lakkaraju, 2020), and several others that find rule-based CEs for single instances (Guidotti et al., 2019; Fern´andez et al., 2020).”
MethodExplanations Go Linear: Post-Hoc Explainability for Tabular Data with Interpretable Meta-Encoding
Piaggesi, S., Guidotti, R., Giannotti, F., & Pedreschi, D. (2025) · Proceedings · doi:10.1109/icdm65498.2025.00074 · Cited by 4
Self-citationOn the complementary roles of XVARS and SHAP in explainable AI
Panigrahi, B., Razavi, S., Doig, L. E., Gupta, H., & Liber, K. (2026) · Environmental Modelling & Software · doi:10.1016/j.envsoft.2026.107007 · Cited by 4
Improving local interpretable classifier explanations exploiting self-generated semantic features
Angiulli, F., Fassetti, F., & Nisticó, S. (2025) · Neural computing & applications (Print) · doi:10.1007/s00521-025-11227-0 · Cited by 3
Robustness of Counterfactual Explanations
Leofante, F., & Wicker, M. (2025) · SpringerBriefs in intelligent systems · doi:10.1007/978-3-031-89022-2_3 · Cited by 3
Explaining Black-Box Clustering Pipelines with Cluster-Explorer
Ofek, S., & Somech, A. (2025) · Proceedings of the VLDB Endowment · doi:10.14778/3718057.3718075 · Cited by 4
“Similar measures to (1) and (2) have been previously proposed in the context of evaluating decision rules [27, 48], drawing parallels to the notions of precision and recall commonly used in supervised learning.”
Causal CSSE: integrating counterfactuals and causality in the explanation of machine learning models
Krauss, O. F. P., Balbino, M. D. S., & Nobre, C. N. (2025) · Machine Learning · doi:10.1007/s10994-025-06856-4 · Cited by 4
“Balbino et al. ( 2023 ) compared CSSE with other counterfactual methods such as the approaches of Wachter et al. ( 2017 ), Dhurandhar et al. ( 2018 ), Guidotti et al. ( 2019 ), Rathi ( 2019 ), Gomez et al. ( 2020 ) and Mothilal et al. ( 2020 ).”
Interpretable Credit Risk Assessment Based on Heuristic Knowledge Extraction Method
Xiao, Z., & Jiao, J. (2021) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai52525.2021.00196 · Cited by 4
Information-seeking dialogue for explainable artificial intelligence: Modelling and analytics
Stepin, I., Budzyńska, K., Catalá, A., Pereira-Fariña, M., & Alonso, J. M. (2023) · Argument & Computation · doi:10.3233/aac-220011 · Cited by 4
Reason to Explain: Interactive Contrastive Explanations (REASONX)
State, L., Ruggieri, S., & Turini, F. (2023) · Communications in computer and information science · doi:10.1007/978-3-031-44064-9_22 · Cited by 4
Self-citationResearch on bronze wine vessel classification using improved SSA-CBAM-GNNs
Wang, W., Miao, S., & Liao, Y. (2024) · PLoS ONE · doi:10.1371/journal.pone.0295690 · Cited by 4
“This may lead to over-fitting issues in classification tasks [18, 19].”
BackgroundIjuice: integer JUstIfied counterfactual explanations
Kuratomi, A., Miliou, I., Lee, Z., Lindgren, T., & Papapetrou, P. (2024) · Machine Learning · doi:10.1007/s10994-024-06530-1 · Cited by 4
“Another tree-based counterfactual generator is the LOcal Rule-based Explainer (LORE) (Guidotti et al., 2019; Guidotti, 2022) which is a local surrogate model that builds a tree on synthetic data to provide both feature relevance and counterfactual rules.”
Yue, J., Wang, J., Zhang, S., Ma, Z., Shi, Y., & Lin, Y. (2024) · Neural Networks · doi:10.1016/j.neunet.2024.106896 · Cited by 4
Guo, M., Zhang, Q., & Zeng, D. D. (2025) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3715150 · Cited by 4
“Some alternative forms of local explanations include providing factual and counterfactual statements [21] and analyzing the nearest neighbors of an instance [30].”
Large Language Models-Based Local Explanations of Text Classifiers
Angiulli, F., Luca, F. D., Fassetti, F., & Nisticò, S. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-78977-9_2 · Cited by 4
Caligiore, D., Monreale, A., Rossetti, G., Bongiorno, A., & Fisicaro, G. (2025) · Frontiers in Complex Systems · doi:10.3389/fcpxs.2025.1604132 · Cited by 4
Self-citationTiainen, M. (2021) · UTUPub (University of Turku) · Cited by 4
Sampling The “Inverse Set” of a Neuron
Hada, S. S., & Carreira-Perpiñán, M. Á. (2021) · Proceedings - International Conference on Image Processing · doi:10.1109/icip42928.2021.9506500 · Cited by 4
“Methods like [4, 5, 6] approximate the deep network locally with a much simple model like decision trees or linear model; and using these simple models, they provide an explanation in terms of input segments (like superpixels in an image).”
MethodDecision Tree Induction Through Meta-learning
Ferreira, C. A., Cantão, A. H., & Baranauskas, J. A. (2022) · IFIP advances in information and communication technology · doi:10.1007/978-3-031-08337-2_9 · Cited by 4
A Modularized Framework for Explaining Black Box Classifiers for Text Data
Sarvmaili, M., Guidotti, R., Monreale, A., Soares, A., Sadeghi, Z., Giannotti, F., Pedreschi, D., & Matwin, S. (2022) · doi:10.21428/594757db.7e75dcdf · Cited by 4
Self-citationConceptualizing the Relationship between AI Explanations and User AgencyPreprint
Adenuga, I. J., & Dodge, J. E. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2312.03193 · Cited by 4
“There are two main approaches employed by post-hoc techniques: opaque box (operates on the input/output boundary; e.g., LIME [35], LORE [20]) and transparent box (operates on the internal structures; e.g., deconvnet model method [49] and network dissection [7, 8]).”
MethodGenerate Counterfactual Explanations for Graph Neural Networks from Node Feature PerturbationsPreprint
Giorgi, F., Silvestri, F., & Tolomei, G. (2024) · doi:10.36227/techrxiv.172866006.66912846/v1 · Cited by 4
Cau, F. M., Desolda, G., Greco, F., Spano, L. D., & Viganò, L. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2507.07916 · Cited by 4
“In the last few years, researchers working in Human-Centred Security and Explainable AI (XAI) have investigated whether feature-based explanations might be effective in mitigating phishing attacks [Guidotti et al., 2019b, Gunning and Aha, 2019].”
Local post-hoc interpretability for black-box classifiers
Laugel, T. (2020) · theses.fr (ABES) · doi:10.70675/afa1c80bz1e1fz4930zb0eaz215e7bb5bf65 · Cited by 4
Johnson, D. K. N. (2025) · The Review of Black Political Economy · doi:10.1177/00346446251331615 · Cited by 3
“Model transparency and interpretability are therefore critical (Guidotti et al., 2019), so that testing can be done of the marginal impacts of inputs on outputs with applications to fairness problems (Adebayo, 2016; Datta et al., 2016; Kilbertus et al., 2017).”
MAINLE: A Multi-Agent, Interactive, Natural Language Local Explainer of Classification Tasks
Serafim, P. B. S., Filho, R. F. F., Freitas, S., Gezici, G., Giannotti, F., Raimondi, F., & dos Santos, A. H. R. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-06078-5_9 · Cited by 3
Self-citationCounterfactual Explanations for Enhanced Interpretability in Cross-Site Scripting (XSS) Detection
Mathew, A. K., Ashwin, M., Kolil, V. K., & Rajeev, D. (2025) · International Conference on Innovative Data Communication Technologies and Application (ICIDCA) · doi:10.1109/icidca66325.2025.11280340 · Cited by 3
“Riccardo et al. [19] described how LIME perturbs input data to approximate a classifier’s decision boundary using a sparse, interpretable model.”
Explainable AI in the Medical Field: A Survey on Machine Learning Interpretability and Use Cases
Thamer, M., & Sultani, Z. N. (2025) · Journal of Al-Nahrain University-Science · doi:10.22401/anjs.28.4.15 · Cited by 3
Hammad, K. M., Salimon, A. I., Statnik, E. S., & Korsunsky, A. M. (2026) · Journal of Materials Research and Technology · doi:10.1016/j.jmrt.2026.01.134 · Cited by 3
DeltaExplainer: A Software Debugging Approach to Generating Counterfactual Explanations
Shree, S., Chandrasekaran, J., Lei, Y., Kacker, R. N., & Kuhn, D. R. (2022) · IEEE International Conference on Artificial Intelligence Testing (AITest) · doi:10.1109/aitest55621.2022.00023 · Cited by 3
“factuals by building a local decision tree in the neighborhood of the original instance using a genetic algorithm, followed by exploring the path of the decision tree [4].”
BackgroundMethodLocal Rule-Based Explanations Method Based on Genetic Algorithms with Fitness Sharing
Santos, D. A., Baranauskas, J. A., & Tinós, R. (2023) · Learning and Nonlinear Models · doi:10.21528/lnlm-vol21-no2-art1 · Cited by 3
Human Centered Approaches and Taxonomies for Explainable Artificial Intelligence
Sheridan, H., Murphy, E., & O’Sullivan, D. M. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-76827-9_9 · Cited by 3
FIPER: A Visual-Based Explanation Combining Rules and Feature Importance
Cappuccio, E., Fadda, D., Lanzilotti, R., & Rinzivillo, S. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-74633-8_11 · Cited by 3
Computational Argumentation and Automatic Rule-Generation for Explainable Data-Driven Modeling
Longo, L., Berretta, S., Verda, D., & Rizzo, L. (2025) · IEEE Access · doi:10.1109/access.2025.3618992 · Cited by 3
“Rule-based interaction supports the production of contrastive explanations, often implemented with the notion of counterfactuals [15], [16], [17], [18].”
The Ethical Impact Assessment of Selling Life Insurance to Titanic Passengers
Gezici, G., Mannari, C., & Orlandi, L. (2023) · Scuola Normale Superiore di Pisa · Cited by 3
Using the K-associated Optimal Graph to Provide Counterfactual Explanations
da Silva, A. T., & Bertini, J. R. (2022) · IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) · doi:10.1109/fuzz-ieee55066.2022.9882751 · Cited by 3
“Guidotti et al [18] proposed LORE (LOcal Rule-based Ex-planation) as a factual and counterfactual explanation method.”
MethodExplainability of Image Classifiers for Targeted Adversarial Attack
Pandya, M. A., Siddalingaswamy, P. C., & Singh, S. (2022) · IEEE India Conference · doi:10.1109/indicon56171.2022.10039871 · Cited by 3
“Guidotti et al. [9] introduced the concept of a model agnostic approach called Local Rule-based Explanations (LORE), which utilizes decision rules formulated using decision trees.”
MethodEvaluation of methods of counterfactual explanation - A qualitative and quantitative analysis
Krauss, O. F. D. P. E., Balbino, M. D. S., & Nobre, C. N. (2023) · doi:10.5753/kdmile.2023.232932 · Cited by 3
Generating Survival Interpretable Trajectories and Data
Konstantinov, A. V., Kirpichenko, S. R., & Utkin, L. V. (2024) · Doklady Mathematics · doi:10.1134/s1064562424601999 · Cited by 3
“A quite different generative model handling survival data and called SurvivalGAN was proposed by Norcliffe et al. [15].”
Sampling the "Inverse Set" of a Neuron: An Approach to Understanding Neural NetsPreprint
Hada, S. S., & Carreira-Perpiñán, M. Á. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1910.04857 · Cited by 3
“Methods like [4, 5, 6] approximate the deep network locally with a much simple model like decision trees or linear model; and using these simple models, they provide an explanation in terms of input segments (like superpixels in an image).”
MethodSpArX: Sparse Argumentative Explanations for Neural Networks [Technical Report]Preprint
Ayoobi, H., Potyka, N., & Toni, F. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2301.09559 · Cited by 3
Deterministic Local Interpretable Model-Agnostic Explanations for Stable Explainability
Zafar, M. R., & Khan, N. (2023) · doi:10.32920/22734320.v1 · Cited by 3
Are numerical or verbal explanations of AI the key to appropriate user reliance and error detection?
Papenkordt, J., Ngonga Ngomo, A.-C., & Thommes, K. (2026) · Behaviour & Information Technology · doi:10.1080/0144929x.2025.2568928 · Cited by 2
Counter-CAM : An Improved Grad-CAM based Visual Explainer for Infrared Breast cancer Classification
Raghavan, K., Sivaselvan, B., & Kamakoti, V. (2023) · IEEE India Conference · doi:10.1109/indicon59947.2023.10440898 · Cited by 2
Handling Veracity of SVM Predictions
Loor, M., Tapia-Rosero, A., & Tré, G. D. (2024) · Lecture notes in networks and systems · doi:10.1007/978-3-031-74003-9_5 · Cited by 2
Outcome-oriented local explanation using context-aware process patterns
Vazifehdoostirani, M., Buliga, A., Genga, L., Francescomarino, C. D., Ronzani, M., Ghidini, C., & Dijkman, R. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.111387 · Cited by 2
Zou, Q., Yang, Z., Qiu, X.-R., Yu, J., Shi, Y., & Chen, N. (2025) · IEEE Transactions on Reliability · doi:10.1109/tr.2025.3611079 · Cited by 2
“Approach: In our comparative analysis, we evaluated the CCE-SDP method against the established baseline methods, specifically CSSE [48], LORE [49], WACH [50], and DiCE [51].”
Toward A Causal Framework for Modeling Perception
Álvarez, J. M., & Ruggieri, S. (2025) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1609/aies.v8i1.36539 · Cited by 2
“In explainable artificial intelligence (xAI), for example, the goal is to develop methods that explain to an expert the predictions of a ML model (Guidotti et al. 2019).”
Self-citation(2025) · doi:10.32620/pi.25.t3 · Cited by 2
Ouartani, S., & Taleb, N. (2025) · Foundations of Computing and Decision Sciences · doi:10.2478/fcds-2025-0019 · Cited by 2
Proximity‐Constrained Counterfactuals for Explainable Diabetes Risk Assessment
Kabir, M. F., Mahakal, A., Wang, Y.-C., Alsobeh, A., Al-Ahmad, B. I., & Alkhawaldeh, R. S. (2025) · Applied Computational Intelligence and Soft Computing · doi:10.1155/acis/3424976 · Cited by 2
Meta‐Explainers: A Unified Ensemble Approach for Multifaceted XAI
Bello, M., Amador, R., García, M.-M., Bello, R., Cordón, Ó., & Herrera, F. (2025) · International Journal of Intelligent Systems · doi:10.1155/int/4841666 · Cited by 2
Metadata
- Authors
- Guidotti, R., Monreale, A., Giannotti, F., Pedreschi, D., Ruggieri, S., & Turini, F.
- Affiliations
- Istituto di Scienza e Tecnologie dell'Informazione "Alessandro Faedo" (IT)
- University of Pisa (IT)
- ORCID
- Venue
- IEEE Intelligent Systems
- Published
- 2019-11-01
- Volume
- 34
- Issue
- 6
- Pages
- 14-23
- ISSN
- 1541-1672 (ISSN-L)
- 1941-1294
- OpenAlex
- W2994120362
- Cited by
- 349
- References
- 14
- Citations by year
- 2026: 38 · 2025: 67 · 2024: 62 · 2023: 60 · 2022: 47 · 2021: 48 · 2020: 26 · 2019: 1
- FWCI
- 13.92
- Type
- journal article
- Language
- en
- Open access
- green
- ieeexplore.ieee.org
- Licence
- Locations
- IEEE Intelligent Systems (publishedVersion)
- CINECA IRIS Institutial research information system (University of Pisa) (submittedVersion, other-oa)
- ISTI Open Portal (submittedVersion, other-oa)
- Scuola Normale Superiore di Pisa (submittedVersion)
- Funders
- European Commission (INFRAIA-1-2014-2015)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Software › Software Testing and Debugging Techniques
- Physical Sciences › Computer Science › Information Systems › Data Mining Algorithms and Applications
- Keywords
- explainable AI, black-box classifiers, counterfactual explanations, genetic algorithm, interpretable classifiers, model transparency, trust in AI
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.