Machines don't actually have bias. AI doesn't'want' something to be true or false for reasons that can't be explained through logic. Unfortunately human bias exists in machine learning from the creation of an algorithm to the interpretation of data – and until now hardly anyone has tried to solve this huge problem. A team of scientists from Czech Republic and Germany recently conducted research to determine the effect human cognitive bias has on interpreting the output used to create machine learning rules. The team's white paper explains how 20 different cognitive biases could potentially alter the development of machine learning rules and proposes methods for "debiasing" them. Biases such as "confirmation bias" (when a person accepts a result because it confirms a previous belief) or "availability bias" (placing greater emphasis on information relevant to the individual than equally valuable information of less familiarity) can render the interpretation of machine learning data pointless.

This paper investigates to what extent do cognitive biases affect human understanding of interpretable machine learning models, in particular of rules discovered from data. Twenty cognitive biases (illusions, effects) are covered, as are possibly effective debiasing techniques that can be adopted by designers of machine learning algorithms and software. While there seems no universal approach for eliminating all the identified cognitive biases, it follows from our analysis that the effect of most biases can be ameliorated by making rule-based models more concise. Due to lack of previous research, our review transfers general results obtained in cognitive psychology to the domain of machine learning. It needs to be succeeded by empirical studies specifically aimed at the machine learning domain.

This paper focuses on care support knowledge (especially focuses on the sleep related knowledge) and tackles its cognitive bias and humanity aspects from machine learning perspective through discussion of whether machine learning can correct commonly accepted knowledge and provide understandable knowledge in care support domain. For this purpose, this paper starts by introducing our data mining method (based on association rule learning) that can provide only necessary number of understandable knowledge without probabilities even if its accuracy slightly becomes worse, and shows its effectiveness in care plans support systems for aged persons as one of healthcare systems. The experimental result indicates that (1) our method can extract a few simple knowledge as understandable knowledge that clarifies what kinds of activities (e.g., rehabilitation, bathing) in care house contribute to having a deep sleep, but (2) the apriori algorithm as one of major association rule learning methods is hard to provide such knowledge because it needs calculate all combinations of activities executed by aged persons.

In this AAAI Spring symposium 2018, we discuss cognitive bias and humanity in the context of well-being AI. We define “well-being AI” as an AI research paradigm for promoting psychological well-being and maximizing human potential. The goals of well-being AI are (1) to understand how our digital experience affects our health and our quality of life and (2) to design well-being systems that put humans at the center. The important challenges of this research are how to quantify subjective things such as happiness, personal impressions, and personal values, and how to transform them into scientific representations with corresponding computational methods. One of the important keywords in understanding machine intelligence in human health and wellness is cognitive bias. Advances in big data and machine learning should not overlook some new threats to enlightened thought, such as the recent trend of social media platforms and commercial recommendation systems being used to manipulate people's inherent cognitive bias. The second important keyword is humanity. Rational thinking, on which early AI researchers had been focused their efforts, is recently and rapidly replacing human thinking by machines. Many people might have begun to believe that irrational thinking is the root of humanity. Empirical and philosophical discussions on AI and humanity would be welcome. This paper describes the detailed motivation, technical, and philosophical challenges of this symposium proposal.

The present approach targets to assist decision-making by identifying and by-passing Cognitive Bias of speakers-participants and evaluators in regard to spoken texts.

Four months ago I attempted to synthesize Wikipedia's crazy list of cognitive biases, and after banging my head against the wall for weeks, came up with this Cognitive Bias Cheat Sheet which John Manoogian III,beautifully organized into the above poster. Since then, I've started working on a book proposal (get on the email list!) around these topics, and wanted to start by creating an actual cheat sheet that doesn't take so long to read. There are four qualities of the universe that limit our own intelligence and the intelligence of every other person, collective, organism, machine, alien, or imaginable god. All 200-ish of our known biases are attempts to work around these conundrums! The first conundrum is that there's too much information in the universe for any individual within the universe to process.

Stay tuned for additional content in this series. So you want to build a cognitive application, but you want it to be great. You want it to be useful, exciting, and inspiring -- in essence, to create a truly cognitive experience. You might be wondering what is a cognitive experience? Should the application I'm designing be cognitive?

When describing routes not in the current environment, a common strategy is to anchor the description in configurations of salient landmarks, complementing the verbal descriptions by "placing" the non-visible landmarks in the gesture space.  Understanding such multimodal descriptions and later locating the landmarks from real world is a challenging task for the hearer, who must interpret speech and gestures in parallel, fuse information from both modalities, build a mental representation of the description, and ground the knowledge to real world landmarks.  In this paper, we model the hearer's task, using a multimodal spatial description corpus we collected.  To reduce the variability of verbal descriptions, we simplified the setup to use simple objects as landmarks.  We describe a real-time system to  evaluate the separate and joint contribution of the modalities. We show that gestures not only help to improve the overall system performance, even if to a large extent they encode redundant information, but also result in earlier final correct interpretations. Being able to build and apply representations incrementally will be of use in more dialogical settings, we argue, where it can enable immediate clarification in cases of mismatch.

Human action recognition remains a difficult problem for AI. Traditional machine learning techniques can have high recognition accuracy, but they are typically black boxes whose internal models are not inspectable and whose results are not explainable. This paper describes a new pipeline for recognizing human actions from skeleton data via analogical generalization. Specifically, starting with Kinect data, we segment each human action by temporal regions where the motion is qualitatively uniform, creating a sketch graph that provides a form of qualitative representation of the behavior that is easy to visualize. Models are learned from sketch graphs via analogical generalization, which are then used for classification via analogical retrieval. The retrieval process also produces links between the new example and components of the model that provide explanations. To improve recognition accuracy, we implement dynamic feature selection to pick reasonable relational features. We show the explanation advantage of our approach by example, and results on three public datasets illustrate its utility.

The Leiter International Performance Scale-Revised (Leiter-R) is a standardized cognitive test that seeks to "provide a nonverbal measure of general intelligence by sampling a wide variety of functions from memory to nonverbal reasoning." Understanding the computational building blocks of nonverbal cognition, as measured by the Leiter-R, is an important step towards understanding human nonverbal cognition, especially with respect to typical and atypical trajectories of child development. One subtest of the Leiter-R, Form Completion, involves synthesizing and localizing a visual figure from its constituent slices. Form Completion poses an interesting nonverbal problem that seems to combine several aspects of visual memory, mental rotation, and visual search. We describe a new computational cognitive model that addresses Form Completion using a novel, mental-rotation-friendly image representation that we call the Polar Augmented Resolution (PolAR) Picture, which enables high-fidelity mental rotation operations. We present preliminary results using actual Leiter-R test items and discuss directions for future work.