Fairness-Aware Multimodal Machine Learning for Retail Stock Prediction from Sentiment and Market Data

PurposeObjectively evaluating the fairness of the algorithm, exploring in specific scenarios combined with scenario characteristics and constructing the algorithm fairness evaluation index system in specific scenarios.Design/methodology/approachThis paper selects marketing scenarios, and in accordance with the idea of “theory construction-scene feature extraction-enterprise practice,” summarizes the definition and standard of fairness, combs the application link process of marketing algorithms and establishes the fairness evaluation index system of marketing equity allocation algorithms. Taking simulated marketing data as an example, the fairness performance of marketing algorithms in some feature areas is measured, and the effectiveness of the evaluation system proposed in this paper is verified.FindingsThe study reached the following conclusions: (1) Different fairness evaluation criteria have different emphases, and may produce different results. Therefore, different fairness definitions and standards should be selected in different fields according to the characteristics of the scene. (2) The fairness of the marketing equity distribution algorithm can be measured from three aspects: marketing coverage, marketing intensity and marketing frequency. Specifically, for the fairness of coverage, two standards of equal opportunity and different misjudgment rates are selected, and the standard of group fairness is selected for intensity and frequency. (3) For different characteristic fields, different degrees of fairness restrictions should be imposed, and the interpretation of their calculation results and the means of subsequent intervention should also be different according to the marketing objectives and industry characteristics.Research limitations/implicationsFirst of all, the fairness sensitivity of different feature fields is different, but this paper does not classify the importance of feature fields. In the future, we can build a classification table of sensitive attributes according to the importance of sensitive attributes to give different evaluation and protection priorities. Second, in this paper, only one set of marketing data simulation data is selected to measure the overall algorithm fairness, after which multiple sets of marketing campaigns can be measured and compared to reflect the long-term performance of marketing algorithm fairness. Third, this paper does not continue to explore interventions and measures to improve algorithmic fairness. Different feature fields should be subject to different degrees of fairness constraints, and therefore their subsequent interventions should be different, which needs to be continued to be explored in future research.Practical implicationsThis paper combines the specific features of marketing scenarios and selects appropriate fairness evaluation criteria to build an index system for fairness evaluation of marketing algorithms, which provides a reference for assessing and managing the fairness of marketing algorithms.Social implicationsAlgorithm governance and algorithmic fairness are very important issues in the era of artificial intelligence, and the construction of the algorithmic fairness evaluation index system in marketing scenarios in this paper lays a safe foundation for the application of AI algorithms and technologies in marketing scenarios, provides tools and means of algorithm governance and empowers the promotion of safe, efficient and orderly development of algorithms.Originality/valueIn this paper, firstly, the standards of fairness are comprehensively sorted out, and the difference between different standards and evaluation focuses is clarified, and secondly, focusing on the marketing scenario, combined with its characteristics, key fairness evaluation links are put forward, and different standards are innovatively selected to evaluate the fairness in the process of applying marketing algorithms and to build the corresponding index system, which forms the systematic fairness evaluation tool of marketing algorithms.

Currently, there is uncertainty surrounding the merits of open-source versus proprietary algorithm development. Though justification in favor of each exists, we argue that open-source algorithm development should be the standard in highly consequential contexts that affect people’s lives for reasons of transparency and collaboration, which contribute to greater predictive accuracy and enjoy the additional advantage of cost-effectiveness. To make this case, we focus on criminal sentencing algorithms, as criminal sentencing is highly consequential, and impacts society and individual people. Furthermore, the popularity of this topic has surged in the wake of recent studies uncovering racial bias in proprietary sentencing algorithms among other issues of over-fitting and model complexity. We suggest that these issues are exacerbated by the proprietary and expensive nature of virtually all widely used criminal sentencing algorithms. Upon replicating a major algorithm using real criminal profiles, we fit three penalized regressions and demonstrate an increase in predictive power of these open-source and relatively computationally inexpensive options. The result is a data-driven suggestion that if judges who are making sentencing decisions want to craft appropriate sentences based on a high degree of accuracy and at low costs, then they should be pursuing open-source options.

Artificial Intelligence (AI) has transformed healthcare by enhancing diagnostic accuracy, treatment personalization, and health service efficiency. However, mounting evidence reveals that AI systems can perpetuate or even amplify existing disparities related to race, gender, socioeconomic status, and geographic location. Biases often originate from imbalanced training datasets, flawed algorithm design, and unequal data collection practices. These biases have led to misdiagnoses, unequal resource allocation, and inadequate treatment recommendations, disproportionately affecting marginalized communities. This review explores the roots of algorithmic bias in healthcare AI, analyzing real-world examples such as COVID-19 triage systems and diagnostic tools that underperform in minority populations. It also examines mitigation strategies, including bias-aware data collection, algorithm design techniques, regulatory frameworks, and stakeholder engagement. Successful case studies and future research directions are presented, emphasizing fairness, transparency, and trust in computational medicine. Establishing robust, bias-resilient AI frameworks is critical to achieving equitable health outcomes and reinforcing the ethical foundations of digital health. Keywords: AI bias, health equity, algorithmic fairness, medical AI, healthcare disparities, machine learning, ethical AI, computational medicine.

The new requirements of Metropolitan Area Networks (MANs) have resulted in the emergence of new data transport technology, Resilient Packet Ring (RPR). The issue of fairness is one of the key technologies in RPR. Whereas traditional definitions for fairness are not applicable for RPR, a fairness conception for RPR based upon ingress aggregated flow is presented in this paper. Then fairness algorithms described in the draft of IEEE standard for RPR is analyzed. In order to overcome the defect under condition of unbalanced data stream, an improvement of the fairness algorithm named RPR-RE which adopts the rate estimation is produced to achieve better performance. At last, the simulation is made and the results are given to prove the conclusion that RPR-RE could eliminate the throughput oscillation approximately compared with the algorithm defined in IEEE standard draft under the scenarios with unbalanced flow. Furthermore, RPR-RE algorithm could implement the fairness between the nodes when multiple congested nodes exist in the ring.

A survey on multimodal bidirectional machine learning translation of image and natural language processing

Artificial intelligence (AI) is increasingly being utilized across various fields of medicine, presenting significant potential for the future of healthcare. This review is to systematically outline the current applications of AI in the field of oral health management and to provide an in-depth analysis of the associated challenges and future opportunities. The review was based on a systematic electronic literature search conducted across databases (PubMed, Web of Science, and Scopus) with the keywords including "artificial intelligence," "AI in dentistry," "tele-dentistry," "oral health education," and "oral health management." English-language studies relevant to the application of AI across various aspects of oral health management were included based on independent assessments by two reviewers. We concluded that in the realm of oral health management, AI technology has diverse applications, including oral health education and counseling, monitoring, screening, diagnosis, treatment, follow-up care of oral diseases, and the collection and management of oral health data. By enhancing public awareness of oral health and improving self-management capabilities, AI can increase diagnostic accuracy, facilitate personalized treatments, support tele-dentistry, optimize the allocation of dental resources, and provide early warnings for oral diseases. These advancements collectively contribute to the efficiency and quality of oral health management. While AI demonstrates considerable promise in this field, several challenges remain, including inconsistencies in oral health data, limited availability and accessibility of data, the reliability of AI-driven results, and issues of bias and fairness in AI algorithms. Addressing these challenges is essential to fully harness the transformative potential of AI in oral health management. Oral health management encompasses the comprehensive handling of oral health risk factors in individuals, populations, and communities through a series of measures and activities aimed at maintaining and promoting oral health. The ultimate goal is to achieve the greatest societal benefit in oral health at the lowest possible cost. By addressing challenges such as data consistency, availability, and reliability, as well as issues of bias and fairness in AI algorithms, AI may play a significant role in oral health management. This paper reviews the role of artificial intelligence in the prevention, diagnosis and treatment of oral diseases, providing an important reference for the later application of artificial intelligence in oral health management.

Recent years have seen growing interest in modeling visitor engagement in museums with multimodal learning analytics. In parallel, there has also been growing concern about issues of fairness and encoded bias in machine learning models. In this paper, we investigate bias detection and mitigation techniques to address issues of algorithmic fairness in multimodal models of museum visitor visual attention. We employ slicing analysis using the Absolute Between-ROC Area (ABROCA) statistic to detect encoded bias present in multimodal models of visitor visual attention trained with facial expression and posture data from visitor interactions with a game-based museum exhibit about environmental sustainability. We investigate instances of gender bias that arise between different combinations of modalities across several machine learning techniques. We also measure the effectiveness of two different debiasing strategies—learned fair representations and reweighing—when applied to the trained multimodal visitor attention models. Results indicate that patterns of bias can arise across different modality combinations for the different visitor visual attention models, and there is often an inherent tradeoff between predictive accuracy and ABROCA. Analyses suggest that debiasing strategies tend to be more effective on multimodal models of visitor visual attention than their unimodal counterparts

The interplay between technology and societal norms often reveals a troubling reality: machine learning systems not only reflect existing gender stereotypes but can also amplify and entrench them, making these biases harder to detect and address. This paper adopts an interdisciplinary approach, combining quantitative and qualitative methods with recent technological advancements, such as machine learning techniques for textual analysis and computational linguistics, to offer a new framework for understanding occupational gender bias in machine learning. The study is motivated by persistent gender inequalities in the labour market and rising concerns about gendered algorithmic bias, as outlined in the European Commission's Gender Equality Strategy 2020-2025. Focusing on language translation technologies, the research explores how machine learning may perpetuate or amplify gender stereotypes, aiming to foster more inclusive digital systems aligned with EU strategic goals. More specifically, it investigates occupational gender segregation and its manifestations in various forms of gender bias in machine learning across English, French, and Greek. The study introduces a classification of gender biases in machine learning, providing insights into professional areas needing intervention to address gender imbalances and identifying enduring stereotypical representations in textual data. To support this, statistical analysis is conducted to explore gender variations in occupations over the past thirteen years, using official data and international classifications such as the International Standard Classification of Occupations (ISCO-08). Moreover, gendered occupational distributions are extracted from 200,920 text instances in the three languages, revealing significant discrepancies between official labour statistics and the training data.

Credit ratings are traditionally generated using models that use financial statement data and market data, which is tabular (numeric and categorical). Practitioner and academic models do not include text data. Using an automated approach to combine long-form text from SEC filings with the tabular data, we show how multimodal machine learning using stack ensembling and bagging can generate more accurate rating predictions. This paper demonstrates a methodology to use big data to extend tabular data models, which have been used by the ratings industry for decades, to the class of multimodal machine learning models.

This study examines the family and social consequences associated with using Artificial Intelligence (AI) to manage workplace conflicts within hybrid workplaces. Hybrid workplaces are a new workplace environment, where technology is used in emerging area of inquiry concerning the degree to which AI related stressors bleed across boundaries into family life. Research has primarily focused on either the technological, or organizational aspects of algorithmic management. As such, there exists a dearth of theoretical understanding as to how workplace systems utilizing AI drive employee psychological security and subsequently their family dynamics. Thus, our objective was to understand how transparency regarding AI, fairness in algorithms used by AI, and digital surveillance employed by employers affects hybrid employee's psychological security and ultimately their family dynamics within the UAE. We utilized an explanatory sequential design to utilize both quantitative survey data collected from 420 hybrid workers, along with qualitative thematic analysis of scenario based vignette responses. Our results indicate a stark disconnect; while participants had moderate confidence in AI systems resolving conflicts; they were extremely concerned about issues of fairness, privacy and lack of algorithmic transparency. Most importantly, 63% of respondents indicated that work-related stress experienced due to AI mediated workplace surveillance and evaluation of performance had a negative effect on communication and emotional stability with family members at home. Therefore, we have proposed the Social Algorithmic Justice Framework (SAJF) to identify the pathways from algorithmic transparency to family resiliency with the influence of social support. Ultimately, this study will provide a theoretically supported framework for integrating workplace AI with socially sustainable values so as to ensure that workplace AI integration enhances family harmony rather than diminishes it.

Chronic myeloid leukemia (CML) has become a paradigm for targeted therapy with BCR-ABL1 tyrosine kinase inhibitors (TKIs). However, the growing volume and complexity of clinical, molecular, and imaging data challenge traditional decision-making based on static risk scores. Digital health technologies and artificial intelligence (AI) offer new opportunities to enhance diagnosis, risk stratification, and treatment personalization in CML.This narrative review is based on a focused literature search in PubMed/MEDLINE and Web of Science (2010-2025), combined with expert selection of pivotal studies in CML, digital health, and AI. We included peer-reviewed original research and reviews describing applications of digitalization, AI, or machine learning (ML) in CML or closely related hematologic malignancies, as well as key publications on ethics, regulation, patient perspectives, and ML operations (MLDevOps).We summarize the integration of electronic health records, telemedicine, networked registries, and real-world evidence as a foundation for AI in CML. We review AI/ML applications in diagnostic hematology (cytomorphology, flow cytometry, cytogenetics, histopathology), prognostic modeling, molecular response monitoring (including automated BCR-ABL1 trend analysis and ghost cytometry), drug discovery, and clinical decision support systems (CDSS). Multimodal ML frameworks that integrate clinical, imaging, histopathological, and genomic data enable more precise disease classification and outcome prediction. At the same time, we discuss challenges related to data quality, algorithmic bias, model transparency, regulatory oversight, and patient trust, emphasizing the need for robust validation and MLDevOps infrastructure.AI has the potential to substantially improve CML diagnosis, prognostication, and treatment selection and to support innovative approaches such as treatment-free remission and rational drug design. However, technical sophistication alone is insufficient. Safe and effective clinical integration of AI in CML will require rigorous multicenter validation, continuous performance monitoring, explainable models aligned with ELN guidelines, appropriate regulatory frameworks, and patient-centered implementation strategies. Under these conditions, AI can become a key enabler of truly personalized, evidence-based, and patient-centered care in CML.

“Big data” is always collected from different resources that have different data structures. With the rapid development of information technologies, current precious data resources are characteristic of multimodes. As a result, based on classical machine learning strategies, multi-modal learning has become a valuable research topic, enabling computers to process and understand “big data”. The cognitive processes of humans involve perception through different sense organs. Signals from eyes, ears, the nose, and hands (tactile sense) constitute a person’s understanding of a special scene or the world as a whole. It reasonable to believe that multi-modal methods involving a higher ability to process complex heterogeneous data can further promote the progress of information technologies. The concepts of multimodality stemmed from psychology and pedagogy from hundreds of years ago and have been popular in computer science during the past decade. In contrast to the concept of “media”, a “mode” is a more fine-grained concept that is associated with a typical data source or data form. The effective utilization of multi-modal data can aid a computer understand a specific environment in a more holistic way. In this context, we first introduced the definition and main tasks of multi-modal learning. Based on this information, the mechanism and origin of multi-modal machine learning were then briefly introduced. Subsequently, statistical learning methods and deep learning methods for multi-modal tasks were comprehensively summarized. We also introduced the main styles of data fusion in multi-modal perception tasks, including feature representation, shared mapping, and co-training. Additionally, novel adversarial learning strategies for cross-modal matching or generation were reviewed. The main methods for multi-modal learning were outlined in this paper with a focus on future research issues in this field.

A modality is an event or experience. Life is multimodal, see, hear, smell, feel, and taste. Multimodal experiences involve some world modalities. Artificial intelligence must grasp multimodal views to understand our surroundings. Multimodal machine learning models interact and correlate input from several modalities. It's a multi-disciplinary field with great potential. In this study, we analyze emerging multimodal machine learning technologies and categorize them scientifically rather than focusing on specific multimodal applications. Multimodal machine learning offers more potential and problems than classifications. Most multimodal learning research collects quantitative data from polls and surveys. This research reviews a detailed library of observational studies on multimodal data (MMD) skills for human learning using artificial intelligence-powered approaches including Machine Learning and Deep Learning. This research also describes how MMD has improved learning and in what environments. This paper discusses multimodal learning and its ongoing improvements and approaches to improving learning. Finally, future researchers should carefully consider building a system that aligns multimodal aspects with the study and learning plan. These elements could enhance multimodal learning by facilitating theory and practice activities. This research lays the groundwork for multimodal data use in future learning technologies and development.

We are perceiving and communicating with the world in a multisensory manner, where different information sources are sophisticatedly processed and interpreted by separate parts of the human brain to constitute a complex, yet harmonious and unified sensing system. To endow the machines with true intelligence, multimodal machine learning that incorporates data from various sources has become an increasingly popular research area with emerging technical advances in recent years. In this paper, we present a survey on multimodal machine learning from a novel perspective considering not only the purely technical aspects but also the intrinsic nature of different data modalities. We analyze the commonness and uniqueness of each data format mainly ranging from vision, audio, text, and motions, and then present the methodological advancements categorized by the combination of data modalities, such as Vision+Text, with slightly inclined emphasis on the visual data. We investigate the existing literature on multimodal learning from both the representation learning and downstream application levels, and provide an additional comparison in the light of their technical connections with the data nature, e.g., the semantic consistency between image objects and textual descriptions, and the rhythm correspondence between video dance moves and musical beats. We hope that the exploitation of the alignment as well as the existing gap between the intrinsic nature of data modality and the technical designs, will benefit future research studies to better address a specific challenge related to the concrete multimodal task, prompting a unified multimodal machine learning framework closer to a real human intelligence system.

Multimodal machine learning, as a prospective advancement in artificial intelligence, endeavors to emulate the brain's multimodal learning abilities with the objective to enhance interactions with humans. However, this approach requires simultaneous processing of diverse types of data, leading to increased model complexity, longer training times, and higher energy consumption. Multimodal neuromorphic devices have the capability to preprocess spatio‐temporal information from various physical signals into unified electrical signals with high information density, thereby enabling more biologically plausible multimodal learning with low complexity and high energy‐efficiency. Here, this work conducts a comparison between the expression of multimodal machine learning and multimodal neuromorphic computing, followed by an overview of the key characteristics associated with multimodal neuromorphic devices. The bio‐plausible operational principles and the multimodal learning abilities of emerging devices are examined, which are classified into heterogeneous and homogeneous multimodal neuromorphic devices. Subsequently, this work provides a detailed description of the multimodal learning capabilities demonstrated by neuromorphic circuits and their respective applications. Finally, this work highlights the limitations and challenges of multimodal neuromorphic computing in order to hopefully provide insight into potential future research directions.