Rank aggregation using active learning in meta-searching

Learning the true rank ordering among objects by aggregating a set of expert opinion rank order lists is an important and ubiquitous problem in many applications ranging from social choice theory to recommendation systems and search aggregation. We study the problem of unsupervised rank aggregation where no ground truth ordering information in available, neither about the true preference ordering among any set of objects nor about the quality of individual rank lists. Aggregating the often inconsistent and poor quality rank lists in such an unsupervised manner is a highly challenging problem, and standard consensus-based methods fall short in terms of both quality and scalability. In this manuscript we propose a novel framework to bypass these issues by using object attributes to augment the standard rank aggregation framework. We design algorithms that learn joint models on both rank lists and object features to obtain an aggregated rank ordering that is more accurate and robust, and also helps weed out rank lists of dubious validity. We validate our techniques on synthetic datasets where our algorithm is able to estimate the true rank ordering even when the rank lists are corrupted. Experiments on three real datasets, MQ2008, MQ2007 and OHSUMED, show that using object features can result in significant improvement in performance over existing rank aggregation methods that do not use object information. Furthermore, when at least some of the rank lists are of high quality, our methods are able to effectively exploit such information to output an aggregated rank ordering of high accuracy.

There is a wealth of literature to support active learning as a superior pedagogical approach in achieving higher order learning compared to didactic teaching. Academics at Monash University are changing their teaching practices to move away from didactic methods to active learning environments. The university is facilitating this change with the recent construction of a state‐of‐the‐art learning and teaching building where academics will utilise active learning methods to engage students and facilitate deeper learning. However amongst the excitement there may be degrees of anxiety being felt by academics and students alike. Students may feel anxious in new learning environments and excessive anxiety can interfere with a students' engagement and academic achievement. Academics may feel anxious as to what to include in their active learning, how to deliver it to produce optimal learning outcomes and feelings of being a novice in a new learning environment.The aims of this study are to investigate student and academic anxiety in active learning situations. Does anxiety to active learning exist and if so, how can it be combatted?. Biomedical science students (years 1–3) and medical students (years 1–2) were invited to complete a survey to gauge their opinions of active learning and any anxiety that active learning and active learning environments may elicit. Over 60% of all biomedical science students agreed that they enjoyed active learning. However, greater than 55% of students were anxious about how they will perform academically in active learning environments and over 50% were nervous in large active learning spaces. Survey data from medical students showed that only 24% of first year students enjoyed active learning compared to 88% of second years. Over 70% of first year students were anxious about how they would perform academically in active learning environments compared to 34% for second years. A student focus group was conducted to discuss active learning and ways to help students reduce their anxiety to active learning.Academics were invited to complete pre‐ and post‐active learning surveys to gauge if anxiety exists to active learning and active learning environments. Of the respondents, 43% of education‐focused (EF) academics were anxious about preparing and delivering active learning. Of the teaching/research (T/R) academics, 43% were anxious about preparing active learning but only 14% were anxious about the delivery. Most EF academics (93%) were anxious about how the active learning would be received by the students compared to 71% for T/R academics. Of the EF academics, 50% agreed they felt like a novice in a new learning environment and 43% felt anxious about changing their teaching style. Of the T/R academics, 43% felt like a novice and 29% felt anxious about changing their teaching style. When the academics were surveyed again post‐active learning, nearly all academics (> 93% EF and T/R) enjoyed delivering the active learning. The general consensus from academics post‐active learning was they felt less anxious about changing their teaching style and preparing/delivering future active learning.Support or Funding InformationMonash University Faculty of MNHS Learning and Teaching Research GrantThis abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.

NOTE: The first page of text has been automatically extracted and included below in lieu of an abstract Session 2632 Active Learning in Large Lectures Russell D. Meier Kansas State University I. Introduction Active learning is powerful teaching. I have used active learning techniques in undergraduate computer engineering courses at Iowa State University and Kansas State University. I firmly believe that the learner-centered, active learning exercises have enhanced student knowledge of the material. I was challenged by two large lecture courses during the 1997-1998 academic year. Each course had a class size approaching 100 students. Many instructors believe that a large lecture is not an appropriate active learning environment. I strongly disagree. Properly designed exercises with appropriate individual-accountability measurements work very well in large lectures. In each course, the students formed cohesive learning teams. The teams enjoyed the active learning challenge problems and directed class discussion. My role evolved from a lecturer to the role of facilitator and mentor. Facilitating these learner-centered classrooms is the most rewarding experience I have had as a faculty member. However, managing large active learning classrooms is not for the weak-at-heart. Large lectures pose unique questions that must be faced when designing active learning environments. How will the large number of teams be selected? Is it possible to diversify teams in a large lecture when the student population is mostly homogenous? How can activities be completed in large lecture halls? What about the students that refuse to participate? Does active learning work for every student in the large course? Will the shear number of active learning teams foster competition, and is competition desired? What about incomplete teams caused by truancy? Is it possible to guarantee individual-accountability within the large student population? I had to find answers to all of these questions. This paper will discuss my experience designing and facilitating the active learning environments for the large lectures of my computer engineering courses. II. Background I began my teaching career in the Department of Electrical and Computer Engineering at Iowa State University. The computer engineering enrollment was increasing steadily while the computer engineering faculty size remained static. Thus, class size increased dramatically for the faculty members teaching core courses. I was the only instructor of one four credit-hour course on microprocessor systems design (CprE 301) from 1994 through 1998. My CprE 301 lectures met for 50 minutes three times per week and every student attended one three-hour laboratory during the week. The lecture grew from 40 students to nearly 100 students during these years. In 1997, I was also gifted with a second large lecture to be taught each semester; a three credit- hour course entitled "Fundamentals of Computer Engineering (CprE 310)." CprE 310 had always been extremely unpopular because it was perceived by students to be a mathematics

The problem of interpreting or aggregating multiple rankings is common to many real-world applications. Perhaps the simplest and most common approach is a weighted rank aggregation, wherein a (convex) weight is applied to each input ranking and then ordered. This paper describes a new tool for visualizing and displaying ranking information for the weighted rank aggregation method. Traditionally, the aim of rank aggregation is to summarize the information from the input rankings and provide one final ranking that hopefully represents a more accurate or truthful result than any one input ranking. While such an aggregated ranking is, and clearly has been, useful to many applications, it also obscures information. In this paper, we show the wealth of information that is available for the weighted rank aggregation problem due to its structure. We apply weight set decomposition to the set of convex multipliers, study the properties useful for understanding this decomposition, and visualize the indifference regions. This methodology reveals information–that is otherwise collapsed by the aggregated ranking–into a useful, interpretable, and intuitive decision support tool. Included are multiple illustrative examples, along with heuristic and exact algorithms for computing the weight set decomposition.

Staying Active

Background Creating an active student-centred learning environment has been shown to improve motivation to learn and retention of knowledge. Facilitation techniques can be utilised to create this learning environment by motivating learners to engage in self-reflection, self-assessment, problem solving, critical thinking, discussion and collaboration. Aim The aim of this teaching programme was to introduce the concept and advantages of active student-centred learning and demonstrate how to create this learning environment with confidence within the team’s own educational sessions. Methods Two one-hour workshops were delivered at a palliative care multidisciplinary teaching meeting. Each workshop initially focused on the topic of active learning. Eight different facilitation techniques were then introduced over the two sessions. Following the session, feedback was requested to explore previous experience and confidence and the likelihood of creating this learning environment in the future. The facilitation techniques included: First session Post it notes; Quiz; Rounds; Buzz groups. Second session Post it notes; Buzz groups; Rounds; Line ups; Brainstorming; Snow balling; Fish bowl. Results Twenty seven members of the multidisciplinary team (MDT) attended the workshops and seven of those members attended both sessions. Some 50% of the MDT had not received prior education regarding this topic and 70% felt more confident at creating this learning environment in the future. Following the first and second sessions, 85% and 62% of the MDT respectively were likely to use all the facilitation techniques demonstrated. Of those who were not likely to use all facilitation techniques introduced in the second session, the fish bowl technique was least popular due to the role-play aspect. Conclusion One-hour workshops focusing on how to create an active student-centred learning environment can improve confidence and the likelihood of learners creating this environment in their own future educational sessions.

A ctive engagement has been found to maximize learning, enhance personal development, and enrich communication in the classroom. This study examined how participants’ perceptions of engagement – measured by questions from the National Survey of Student Engagement (NSSE) - changed in response to three problem-based learning activities; peer-led workshops, service learning, and journal clubs. Sixty-two undergraduate students enrolled in an introductory class to a career in physical therapy were recruited for three consecutive semesters. Participants were administered a shortened version of the NSSE at the beginning and end of their semester of class. Two NSSE benchmarks - active and collaborative learning and supportive campus environment – were perceived by participants as improved, partially supporting the hypothesis. Participants reported increased engagement in six scalelets (specific subsets of educational experiences) - course challenge, writing, active learning, collaborative learning, diversity, and interpersonal environment . Responses to several scalelets, however, showed a decline in engagement - higher order thinking and information technology. This is the first study to look at changes in NSSE scores across only one semester of class. The study supports problem-based learning, however, recognizes that class design, composition, and instructor may impact student perceptions.

In this work, we study the rank aggregation problem based on election voting model. In addition to classical voting rules: Borda, some elimination voting algorithms, including Kemeny methods, are analyzed in a graph theoretic approach. As the Kemeny ranking problem is NP-hard, a new heuristic elimination voting algorithm is proposed in this work. Some experiments have been carried out on TREC data for evaluating these voting algorithms on rank aggregation. The experiments indicate that these elimination algorithms have comparable performance with Borda algorithms, and sometimes outperform it

Over the past two decades there has been a concerted effort to move accounting education from a knowledge-oriented system to a process-oriented approach. Active learning is an internationally recognized cornerstone of this strategy. A new technology offers opportunities for professors to promote active learning in their classrooms. That technology frequently called student response systems (SRSs) enable students to answer questions posed by an instructor via a remote control device. This study examines the impact that a new technology, SRS, had on students’ perceptions of key active learning characteristics. The results suggest that students receiving an SRS treatment registered significantly stronger agreement with statements describing an active learning environment than their peers in a control group. The results led to the conclusion that SRSs can be used to facilitate active learning in accounting classrooms. The survey also examined the impact of the SRS treatment on student perceptions regarding classroom efficiency. Students receiving the SRS treatment registered stronger agreement with statements suggesting that the learning environment was efficient. In summary, the survey results suggest that SRS technology is a promising teaching tool that deserves further scrutiny by accounting educators.

Supervised learning, e.g., classification, plays an important role in processing and organizing microblogging data. In microblogging, it is easy to mass vast quantities of unlabeled data, but would be costly to obtain labels, which are essential for supervised learning algorithms. In order to reduce the labeling cost, active learning is an effective way to select representative and informative instances to query for labels for improving the learned model. Different from traditional data in which the instances are assumed to be independent and identically distributed (i.i.d.), instances in microblogging are networked with each other. This presents both opportunities and challenges for applying active learning to microblogging data. Inspired by social correlation theories, we investigate whether social relations can help perform effective active learning on networked data. In this paper, we propose a novel Active learning framework for the classification of Networked Texts in microblogging (ActNeT). In particular, we study how to incorporate network information into text content modeling, and design strategies to select the most representative and informative instances from microblogging for labeling by taking advantage of social network structure. Experimental results on Twitter datasets show the benefit of incorporating network information in active learning and that the proposed framework outperforms existing state-of-the-art methods.

An international association advancing the multidisciplinary study of informing systems. Founded in 1998, the Informing Science Institute (ISI) is a global community of academics shaping the future of informing science.

There is considerable evidence from the physics education literature that traditional approaches are ineffective in teaching physics concepts. A better teaching method is to use the active learning environment, which can be created using interactive lecture demonstrations. Based on the active learning methodology and within the framework of the UNESCO mandate in physics education and introductory physics, the ALOP project (active learning in optics and photonics) was started in 2003, to provide a focus on an experimental area that is adaptable and relevant to research and educational conditions in many developing countries. This project is discussed in this paper.

In a variety of instructional scenarios, librarians at The University of North Carolina have been working to engage students with the use of interactive technologies combined with active learning strategies in attempts to enhance learning of information literacy concepts. This paper describes the successful use of this approach in not only increasing student engagement, but possibly learning outcomes as well. Practical examples of our use of class response systems (clickers), interactive whiteboards, wireless slates, and digital cameras in conjunction with active learning environments are discussed in the context of both upper and lower-division classes. Our study, using a clicker group and control group in information literacy sessions for English Composition II classes, revealed a four-point increase in information retention in classes using clickers and discussion compared to classes taught using more traditional methods of instruction. Assessment data collected from experiments provides evidence that these technologies coupled with active learning methods are capable of delivering not only improved student engagement and participation, but also increased learning outcomes in some cases. The results of these experiences shed light on the potential impact that implementation of interactive technologies with a shift in pedagogy can have on student engagement and learning of information literacy concepts.

Web document ranking via active learning and kernel principal component analysis

Tackling the rank aggregation problem with evolutionary algorithms