More Like this

1. A Ranked Bandit Approach for Multi-stakeholder Recommender Systems

   Arabghalizi, Tahereh ; Labrinidis, Alexandros ( September 2022 , Workshop of Multi-Objective Recommender Systems (MORS’22))

   Recommender systems traditionally find the most relevant products or services for users tailored to their needs or interests but they ignore the interests of the other sides of the market (aka stakeholders). In this paper, we propose to use a Ranked Bandit approach for an online multi-stakeholder recommender system that sequentially selects top 𝑘 items according to the relevance and priority of all the involved stakeholders. We presented three different criteria to consider the priority of each stakeholder when evaluating our approach. Our extensive experimental results on a movie dataset showed that the contextual multi-armed bandits with a relevance function make a higher level of satisfaction for all involved stakeholders in the long term. Keywords: Multi-stakeholder Recommender Systems; Multi-armed Bandits; Ranked Bandit; 

   more » « less
2. Integrated SDM database: Enhancing the relevance and utility of species distribution models in conservation management

   https://doi.org/10.1111/2041-210X.13736  

   Frans, Veronica F. ; Augé, Amélie A. ; Fyfe, Jim ; Zhang, Yuqian ; McNally, Nathan ; Edelhoff, Hendrik ; Balkenhol, Niko ; Engler, Jan O. ( November 2021 , Methods in Ecology and Evolution)

   Species' ranges are changing at accelerating rates. Species distribution models (SDMs) are powerful tools that help rangers and decision‐makers prepare for reintroductions, range shifts, reductions and/or expansions by predicting habitat suitability across landscapes. Yet, range‐expanding or ‐shifting species in particular face other challenges that traditional SDM procedures cannot quantify, due to large differences between a species' currently occupied range and potential future range. The realism of SDMs is thus lost and not as useful for conservation management in practice. Here, we address these challenges with an extended assessment of habitat suitability through anintegrated SDM database(iSDMdb).

   TheiSDMdbis a spatial database of predicted sites in a species' prediction range, derived from SDM results, and is a single spatial feature that contains additional, user‐friendly data fields that synthesise and summarise SDM predictions and uncertainty, human impacts, restoration features, novel preferences in novel spaces and management priorities. To illustrate its utility, we used the endangered New Zealand sea lionPhocarctos hookeri. We consulted with wildlife rangers, decision‐makers and sea lion experts to supplement SDM predictions with additional, more realistic and applicable information for management.

   Almost half the data fields included in this database resulted from engaging with these end‐users during our study. The SDM found 395 predicted sites. However, theiSDMdb's additional assessments showed that the actual suitability of most sites (90%) was questionable due to human impacts. >50% of sites contained unnatural barriers (fences, grazing grasslands), and 75% of sites had roads located within the species' range of inland movement. Just 5% of the predicted sites were mostly (>80%) protected.

   Integrating SDM results with supplemental assessments provides a way to address SDM limitations, especially for range‐expanding or ‐shifting species. SDM products for conservation applications have been critiqued for lacking transparency and interpretation support, and ineffectively communicating uncertainty. TheiSDMdbaddresses these issues and enhances the practical relevance and utility of SDMs for stakeholders, rangers and decision‐makers. We exemplify how to build aniSDMdbusing open‐source tools, and how to make diverse, complex assessments more accessible for end‐users.

    

   more » « less
3. Rationally Inattentive Inverse Reinforcement Learning Explains YouTube Commenting Behavior

   Hoiles, William ; Krishnamurthy, Vikram ; Pattanayak, Kunal ( September 2020 , Journal of machine learning research)

   null (Ed.)

   We consider a novel application of inverse reinforcement learning with behavioral economics constraints to model, learn and predict the commenting behavior of YouTube viewers. Each group of users is modeled as a rationally inattentive Bayesian agent which solves a contextual bandit problem. Our methodology integrates three key components. First, to identify distinct commenting patterns, we use deep embedded clustering to estimate framing information (essential extrinsic features) that clusters users into distinct groups. Second, we present an inverse reinforcement learning algorithm that uses Bayesian revealed preferences to test for rationality: does there exist a utility function that rationalizes the given data, and if yes, can it be used to predict commenting behavior? Finally, we impose behavioral economics constraints stemming from rational inattention to characterize the attention span of groups of users. The test imposes a Rényi mutual information cost constraint which impacts how the agent can select attention strategies to maximize their expected utility. After a careful analysis of a massive YouTube dataset, our surprising result is that in most YouTube user groups, the commenting behavior is consistent with optimizing a Bayesian utility with rationally inattentive constraints. The paper also highlights how the rational inattention model can accurately predict commenting behavior. The massive YouTube dataset and analysis used in this paper are available on GitHub and completely reproducible 

   more » « less
4. Engagement in Practice: Lessons Learned Partnering with Science Educators and Local Engineers in Rural Schools

   Lesko, H. L. ; Grohs, J. R. ; Matusovich, H. M. ; Kirk, G. R. ; Carrico, C. ; van Montfrans, V. ; Gillen, A. L. ; Paradise, T. ; Blackowski, S. A. ; Baum, L. M. ( June 2018 , ASEE Annual Conference proceedings)

   Our NSF-funded ITEST project focuses on the collaborative design, implementation, and study of recurrent hands-on engineering activities with middle school youth in three rural communities in or near Appalachia. To achieve this aim, our team of faculty and graduate students partner with school educators and industry experts embedded in students’ local communities to collectively develop curriculum to aim at teacher-identified science standard and facilitate regular in-class interventions throughout the academic year. Leveraging local expertise is especially critical in this project because family pressures, cultural milieu, and preference for local, stable jobs play considerable roles in how Appalachian youth choose possible careers. Our partner communities have voluntarily opted to participate with us in a shared implementation-research program and as our project unfolds we are responsive to community-identified needs and preferences while maintaining the research program’s integrity. Our primary focus has been working to incorporate hands-on activities into science classrooms aimed at state science standards in recognition of the demands placed on teachers to align classroom time with state standards and associated standardized achievement tests. Our focus on serving diverse communities while being attentive to relevant research such as the preference for local, stable jobs attention to cultural relevance led us to reach out to advanced manufacturing facilities based in the target communities in order to enhance the connection students and teachers feel to local engineers. Each manufacturer has committed to designating several employees (engineers) to co-facilitate interventions six times each academic year. Launching our project has involved coordination across stakeholder groups to understand distinct values, goals, strengths and needs. In the first academic year, we are working with 9 different 6th grade science teachers across 7 schools in 3 counties. Co-facilitating in the classroom are representatives from our project team, graduate student volunteers from across the college of engineering, and volunteering engineers from our three industry partners. Developing this multi-stakeholder partnership has involved discussions and approvals across both school systems (e.g., superintendents, STEM coordinators, teachers) and our industry partners (e.g., managers, HR staff, volunteering engineers). The aim of this engagement-in-practice paper is to explore our lessons learned in navigating the day-to-day challenges of (1) developing and facilitating curriculum at the intersection of science standards, hands-on activities, cultural relevancy, and engineering thinking, (2) collaborating with volunteers from our industry partners and within our own college of engineering in order to deliver content in every science class of our 9 6th grade teachers one full school day/month, and (3) adapting to emergent needs that arise due to school and division differences (e.g., logistics of scheduling and curriculum pacing), community differences across our three counties (e.g., available resources in schools), and partner constraints. 

   more » « less
5. Learning Contextual Bandits in a Non-stationary Environment

   https://doi.org/10.1145/3209978.3210051  

   Wu, Qingyun ; Iyer, Naveen ; Wang, Hongning ( July 2018 , SIGIR '18 The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval)

   Multi-armed bandit algorithms have become a reference solution for handling the explore/exploit dilemma in recommender systems, and many other important real-world problems, such as display advertisement. However, such algorithms usually assume a stationary reward distribution, which hardly holds in practice as users' preferences are dynamic. This inevitably costs a recommender system consistent suboptimal performance. In this paper, we consider the situation where the underlying distribution of reward remains unchanged over (possibly short) epochs and shifts at unknown time instants. In accordance, we propose a contextual bandit algorithm that detects possible changes of environment based on its reward estimation confidence and updates its arm selection strategy respectively. Rigorous upper regret bound analysis of the proposed algorithm demonstrates its learning effectiveness in such a non-trivial environment. Extensive empirical evaluations on both synthetic and real-world datasets for recommendation confirm its practical utility in a changing environment. 

   more » « less