Despite substantial progress in applying neural networks (NN) to a wide variety of areas, they still largely suffer from a lack of transparency and interpretability. While recent developments in explainable artificial intelligence attempt to bridge this gap (e.g., by visualizing the correlation between input pixels and final outputs), these approaches are limited to explaining low-level relationships, and crucially, do not provide insights on error correction. In this work, we propose a framework (VRX) to interpret classification NNs with intuitive structural visual concepts. Given a trained classification model, the proposed VRX extracts relevant class-specific visual concepts and organizes them using structural concept graphs (SCG) based on pairwise concept relationships. By means of knowledge distillation, we show VRX can take a step towards mimicking the reasoning process of NNs and provide logical, concept-level explanations for final model decisions. With extensive experiments, we empirically show VRX can meaningfully answer "why" and "why not" questions about the prediction, providing easy-to-understand insights about the reasoning process. We also show that these insights can potentially provide guidance on improving NN's performance.

An essential part of argumentation-based reasoning is to identify arguments in favor and against a statement or query, select the acceptable ones, and then determine whether or not the original statement should be accepted. We present here an abstract framework that considers two independent forms of argument interaction--support and conflict--and is able to represent distinctive information associated with these arguments. This information can enable additional actions such as: (i) a more in-depth analysis of the relations between the arguments; (ii) a representation of the user's posture to help in focusing the argumentative process, optimizing the values of attributes associated with certain arguments; and (iii) an enhancement of the semantics taking advantage of the availability of richer information about argument acceptability. Thus, the classical semantic definitions are enhanced by analyzing a set of postulates they satisfy. Finally, a polynomial-time algorithm to perform the labeling process is introduced, in which the argument interactions are considered.

It is (2017) considered bugs as implementation errors, gaining more and more attention these days since similar to software bugs, while Cadamuro et al. explanations are necessary in several applications, (2016) defined a bug as a particularly damaging especially in high-stake domains such as healthcare, or inexplicable test error. In this paper, we follow law, transportation, and finance (Adadi and the definition of (model) bugs from Adebayo Berrada, 2018). Some researchers have explored et al. (2020) as contamination in the learning and/or various merits of explanations to humans, such as prediction pipeline that makes the model produce supporting human decision makings (Lai and Tan, incorrect predictions or learn error-causing associations.

Explainable AI (XAI) has long been a fringe discipline in the broader world of AI and machine learning. It exists because many machine-learning models are either opaque or so convoluted that they defy human understanding. But why is it such a hot topic today? AI systems making inexplicable decisions are your governance, regulatory, and compliance colleagues' worst nightmare. But aside from this, there are other compelling reasons for shining a light into the inner workings of AI.

Artificial intelligence (AI) has achieved growing momentum in its application in many fields to deal with the increased complexity, scalability, and automation, and that also permeates into digital networks today. A rapid surge in the complexity and sophistication of AI-powered systems has evolved to such an extent that humans do not understand the complex mechanisms by which AI systems work or how they make certain decisions -- something that is particularly a challenge when AI-based systems compute outputs that are unexpected or seemingly unpredictable. This especially holds true for opaque decision- making systems, such as those using deep neural networks (DNNs), which are considered complex black box models. The inability for humans to see inside black boxes can result in AI adoption (and even its further development) being hindered, which is why growing levels of autonomy, complexity, and ambiguity in AI methods continues to increase the need for interpretability, transparency, understandability, and explainability of AI products/outputs (such as predictions, decisions, actions, and recommendations). These elements are crucial to ensuring that humans can understand and -- consequently -- trust AI-based systems (Mujumdar, et al., 2020). Explainable artificial intelligence (XAI) refers to methods and techniques that produce accurate, explainable models of why and how an AI algorithm arrives at a specific decision so that AI solution results can be understood by humans (Barredo Arrieta, et al., 2020).

Artificial intelligence (AI) is becoming increasingly more popular and can be found in workplaces and homes around the world. However, how do we ensure trust in these systems? Regulation changes such as the GDPR mean that users have a right to understand how their data has been processed as well as saved. Therefore if, for example, you are denied a loan you have the right to ask why. This can be hard if the method for working this out uses "black box" machine learning techniques such as neural networks. TrustyAI is a new initiative which looks into explainable artificial intelligence (XAI) solutions to address trustworthiness in ML as well as decision services landscapes. In this paper we will look at how TrustyAI can support trust in decision services and predictive models. We investigate techniques such as LIME, SHAP and counterfactuals, benchmarking both LIME and counterfactual techniques against existing implementations. We also look into an extended version of SHAP, which supports background data selection to be evaluated based on quantitative data and allows for error bounds.

In this paper, we propose a fresh perspective on argumentation semantics, to view them as a relational database. It offers encapsulation of the underlying argumentation graph, and allows us to understand argumentation semantics under a single, relational perspective, leading to the concept of relational argumentation semantics. This is a direction to understand argumentation semantics through a common formal language. We show that many existing semantics such as explanation semantics, multi-agent semantics, and more typical semantics, that have been proposed for specific purposes, are understood in the relational perspective.

Explainable AI provides a whole new layer of insight by allowing analysts to clearly see why a prediction was made. When it comes to Enterprise AI (Artificial Intelligence), we often focus on automating repetitive business processes for a very simple reason and it doesn't take much imagination to see the value. But what if you wanted to gauge the impact of an unexpected event, such as a hurricane, on your business's bottom line? Maybe you'd like to compare the probable financial outcomes of a strategic decision before you make it? Explainable AI, which combines Human Intelligence with Artificial Intelligence, means employees now have the visibility to make these decisions.

We argue that an explainable artificial intelligence must possess a rationale for its decisions, be able to infer the purpose of observed behaviour, and be able to explain its decisions in the context of what its audience understands and intends. To address these issues we present four novel contributions. Firstly, we define an arbitrary task in terms of perceptual states, and discuss two extremes of a domain of possible solutions. Secondly, we define the intensional solution. Optimal by some definitions of intelligence, it describes the purpose of a task. An agent possessed of it has a rationale for its decisions in terms of that purpose, expressed in a perceptual symbol system grounded in hardware. Thirdly, to communicate that rationale requires natural language, a means of encoding and decoding perceptual states. We propose a theory of meaning in which, to acquire language, an agent should model the world a language describes rather than the language itself. If the utterances of humans are of predictive value to the agent's goals, then the agent will imbue those utterances with meaning in terms of its own goals and perceptual states. In the context of Peircean semiotics, a community of agents must share rough approximations of signs, referents and interpretants in order to communicate. Meaning exists only in the context of intent, so to communicate with humans an agent must have comparable experiences and goals. An agent that learns intensional solutions, compelled by objective functions somewhat analogous to human motivators such as hunger and pain, may be capable of explaining its rationale not just in terms of its own intent, but in terms of what its audience understands and intends. It forms some approximation of the perceptual states of humans.

Deep Learning has become a very valuable tool in different fields, and no one doubts the learning capacity of these models. Nevertheless, since Deep Learning models are often seen as black boxes due to their lack of interpretability, there is a general mistrust in their decision-making process. To find a balance between effectiveness and interpretability, Explainable Artificial Intelligence (XAI) is gaining popularity in recent years, and some of the methods within this area are used to generate counterfactual explanations. The process of generating these explanations generally consists of solving an optimization problem for each input to be explained, which is unfeasible when real-time feedback is needed. To speed up this process, some methods have made use of autoencoders to generate instant counterfactual explanations. Recently, a method called Deep Guided Counterfactual Explanations (DGCEx) has been proposed, which trains an autoencoder attached to a classification model, in order to generate straightforward counterfactual explanations. However, this method does not ensure that the generated counterfactual instances are close to the data manifold, so unrealistic counterfactual instances may be generated. To overcome this issue, this paper presents Distribution Aware Deep Guided Counterfactual Explanations (DA-DGCEx), which adds a term to the DGCEx cost function that penalizes out of distribution counterfactual instances.