Narrow Results

Explainability and comprehensibility of AI are important requirements for intelligent systems deployed in real-world domains. Users want and frequently need to understand how decisions impacting them are made. Similarly, it is important to understand how an intelligent system functions for safety and security reasons. In this paper, we describe two complementary impossibility results (Unexplainability and Incomprehensibility), essentially showing that advanced AIs would not be able to accurately explain some of their decisions and for the decisions they could explain people would not understand some of those explanations.

Software metrics based maintainability prediction is leading to development of new sophisticated techniques to construct prediction models. This paper proposes a new software maintainability prediction framework, which bases on Fuzzy Network, a novel exploratory modeling technique. The proposed framework utilizes both the metric data collected from software system and the subjective appraisals from experts. An application example of the framework is shown. In comparison to the Standard Fuzzy System based models, Fuzzy Network based models improves the transparency more than 71.3% and the accuracy more than 11.0%. It is confirmed that Fuzzy Network based framework is more appropriate for constructing SMP model.

Corporate governance (CG) reformists typically presume better-governed companies are more transparent to investors. We focus on CG and transparency in Japan, where CG has been an ongoing issue. Using local ratings of Japanese companies’ CG and data on corporate disclosures and their associated stock returns, we do find better-governed Japanese companies have made more frequent and timelier disclosures, and their share prices have reflected value-relevant information earlier. While these results hold for good news, they do not hold for bad. Consequently, governance guidance in Japan may not have resulted in both timelier and more balanced release of newsworthy information.

It is easier than ever to run modern machine learning (ML) models. However, developing and implementing systems that support real-world risk management applications in a bank is a significant challenge. It is partly because ML models are not transparent and explainable. The framework presented here covers the leading eXplainable AI (XAI) methods. Practical challenges in implementing these methods are discussed.

How an organisation responds to crises and disasters has significant implications, for not only its reputation and brand but also its business continuity. We propose that four important considerations, power, honesty, transparency and timeliness, must be addressed when designing crisis communication. We then discuss how failure to give these priority can lead to emotional contagion, conspiracy theories, moral panics and fake news. We conclude by proposing that if crisis communication management is to be effective, then not only should it be ethical and socially responsible but also contingent, stakeholder centric and involving cycles of sensemaking and sense giving.

A novel methodology is proposed for the development of neural network models for complex engineering systems exhibiting nonlinearity. This method performs neural network modeling by first establishing some fundamental nonlinear functions from a priori engineering knowledge, which are then constructed and coded into appropriate chromosome representations. Given a suitable fitness function, using evolutionary approaches such as genetic algorithms, a population of chromosomes evolves for a certain number of generations to finally produce a neural network model best fitting the system data. The objective is to improve the transparency of the neural networks, i.e. to produce physically meaningful "white box" neural network model with better generalization performance. In this paper, the problem formulation, the neural network configuration, and the associated optimization software are discussed in detail. This methodology is then applied to a practical real-world system to illustrate its effectiveness.

The goal of online examination is to ensure convenience, scalability, expanded reach, and customization. Despite the popularity of online exams, there have been some demonstrations against using them. The existing system is vulnerable and faces failures like question paper leakage, answer deception, trying to forge scorecards, difficulty in ensuring the integrity and consistency of scorecards by the exam board, and so on. Blockchain technology is the greatest and unsurpassed option for continuous authentication, confidentiality, and reliability. A blockchain-based framework for an online examination system is proposed to collect, register, and maintain a growing list of records using cryptography. It is also a peer-to-peer system that manages or stores time-stamped transactions such that no one can alter them. The proposed model combines time lock and multi-signature techniques and ensures the encrypted questions are locked till the exam starts, candidates’ answers are not altered, and scorecards are not manipulated. The system is open to the public, and the system’s processes are transparent. The costs associated are analyzed with the suggested solution’s on-chain and off-chain evaluation.

Many of the artificial intelligence (AI) systems used nowadays have a very high level of accuracy but fail to explain their decisions. This is critical, especially in sensitive areas such as medicine and the health area at large but also for applications of the law, finance etc., where explanations for certain decisions are needed and are often useful and valuable as the decision itself. This paper presents a review of four different methods for creating transparency in AI systems. It also suggests a list of criteria under which circumstances one should use which methods.

We examine the impact of iceberg orders on the price and order flow dynamics in limit order books. Iceberg orders allow traders to simultaneously hide a large portion of their order size and signal their interest in trading to the market. We show that when market participants detect iceberg orders they tend to strongly respond by submitting matching market orders consistent with iceberg orders facilitating the search for latent liquidity. The greater the fraction of an iceberg order that is executed, the smaller is its price impact consistent with liquidity rather than informed trading. The presence of iceberg orders is associated with increased trading consistent with a positive liquidity externality, but the reduced order book transparency associated with iceberg orders also creates an adverse selection cost for limit orders that may partly offset any gains.

The cockroach farming industry in China has grown without much publicity. However, after a building housing a cockroach farm was leveled in the city of Dafeng, Jiangsu province, under the assumption it was vacant, millions of cockroaches escaped out into the open. Authorities had to work quickly to prevent the swarm of roaches from reaching farming and residential areas. This case breaks open the ethical issues involved in cockroach farming and discusses questions of industry transparency.

How does the sovereign credit ratings history provided by independent ratings agencies affect domestic financial sector development and international capital inflows to emerging countries? We address this question utilizing a comprehensive dataset of sovereign credit ratings from Standard and Poor’s from 1995–2003 for a cross-section of 51 emerging markets. Within a panel data estimation framework, we examine financial sector development and the influence of sovereign credit ratings provision, controlling for various economic and corporate governance factors identified in the financial development literature. We find strong evidence that our sovereign credit rating measures do affect financial intermediary sector developments and capital flows. We find that: (i) long-term foreign currency sovereign credit ratings are important for encouraging financial intermediary development and for attracting capital flows; (ii) Long-term local currency ratings stimulate domestic market growth but discourage international capital flows; (iii) Short-term ratings (both foreign and local currency denominated) retard all forms of financial developments and capital flows. There are important implications in this research for policy makers to encourage the provision of longer-term credit ratings to promote financial development in emerging economies.