About this Research Topic
Big data analytics is at the epicenter of an ongoing global-scale digital transformation. One of the primary mainstays of this transformation is the revolutionization of the strategic decision-making process paradigm, shifting the emphasis from intuition and information to knowledge. Among the analytics designed for the latter paradigm stand out those taking into consideration the behavior of agents, namely humans, autonomous software components, IoT devices, or any other digital entity for that matter which can make independent data-driven decisions. Behavioral analytics provide deeper insight into the motives behind past actions and have greater predictive power for future ones. This emerging change has been already fruitful in diverse areas such as social media analysis, business management and risk estimation, online political and commercial campaigns, as well as healthcare system organization.
At the very heart of the class of behavioral analytics lies the ability to augment decision-making models with psychological, cognitive, emotional, cultural, and social aspects in the decision-making process. Given the volume, speed, and multimodality of these data sources, increased computational power is required if meaningful results are to be generated in a time interval reasonable enough for a plan or policy to be made based on them. This power may well come from distributed platforms or high-end GPU clusters along with specialized algorithms which can harness that power. The latter includes signal estimation methods for discovering patterns in facial expressions, video, or voice, deep learning techniques to process low and high-level attributes of social media posts, state-space models to track emotional changes, and recommender systems that can take into consideration emotional states. Considering as an example the case of social media analysis, several behavioral analytics can be set up over them to collect enormous amounts of data in order to develop effective algorithms for agent simulations, brand loyalty, link prediction, community structure discovery, information diffusion, digital influence, sentiment assessment, fake news dissemination and source trace, and bot discovery.
Fundamental concepts from behavioral economics like bounded rationality, perceived risk, preference to default options, personality traits, human influence, environmental factors, attention management, and loss aversion are driving our everyday life and are progressively integrated to create a new form of digital awareness. Behavioral economics has a multidisciplinary nature and does not negate already established findings in the application fields or even in classical economics but rather extends the respective working methodologies in three directions. First, the human decision-making process is seen from a different view with more interacting factors. Second, behavioral analytics operate on a finer granularity since they analyze individuals and groups in addition to large populations. Third, the above factors become time-dependent as individuals respond to new data, interact with each other, or receive feedback on past decisions.
Strategies based on behavioral analytics have found so far several applications such as in education for improving student nutrition plans, banks for reducing the risk of default loans, HR departments for selecting the appropriate candidates for high profile positions, real estate agencies for increasing the interest of potential buyers, legal contracting for automating smart contracts adjusting their creation and execution to the parties involved, and in airports for efficiently moving transit passengers to their connecting flights.
The primary objective of this issue is to explore the recent developments in big data analytics for computational behavior economics and address theoretical and practical challenges in the field. Original research articles are encouraged, especially those of interdisciplinary nature, including but not limited to the following topics:
Adaptive nudge models
Affective state in social media beyond single polarity
Affective state discovery in argumentation graphs in enterprise environments
Affective computing and human emotional feedback
Argumentation graph mining
Behavioral economics and deep learning
Bounded rationality estimation techniques and computational models
Behavior analysis in social networks
Case studies of behavioral economics
Computational affective models
Computational infrastructure for emotional data and related analytics
Control strategies for state-space affective models
Data-driven approaches for strategy recommendation
Deep learning strategies for training nudge theory models
Discovering emotion-based communities in static and temporal graphs
Distributed approaches to computing public sentiment at a massive scale
Gamified strategies and possibly the results of their application
Graph neural networks for assessing collective and node sentiment in social graphs
Higher-order sentiment metrics with heterogeneous factors
Identifying influential accounts with emotion augmented metrics
Information augmentation through social media and crowdsourcing
Loss aversion strategy discovery with deep learning models
Mining emotion sequences of social media accounts
Natural language processing for emotional aspect mining
Nudge theory-based strategies for brand loyalty and social media campaigns
Ontologies for affective state estimation
Risk estimators based on perceptive criteria and incomplete information
Semantics for Barnum statements and methodologies for their discovery
Smart agent social activity and training
Social software components training and adaptive collaboration schemes
State-space models for emotion changes
The proposed collection is expected to become a focal point for researchers and practitioners of diverse backgrounds by collecting high-quality articles. Emphasis will be placed on case studies, datasets, and information systems with behavioral components. Moreover, it is anticipated that the extension of computer science methodologies to behavioral economics will attract authors from the former field and create new datasets, possibly multimodal or highly heterogeneous. It is expected that this Research Topic will provide a forum where research related to big data analytics for behavioral economics will be seen as a new emerging multidisciplinary research area. To further enhance that point, this Frontiers volume is launched in conjunction with the THECOG workshop, which this year will be part of CIKM 2021. Authors of 10 selected papers from THECOG are invited to submit extended versions of their respective work at a special discount rate.
Keywords: Big data analytics, decision-making models, computational behavioral economics
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.