Much work in knowledge extraction from text tacitly assumes that the frequency with which people write about actions, outcomes, or properties is a reflection of real-world frequencies or the degree to which a property is characteristic of a class of individuals. In this paper, we question this idea, examining the phenomenon of reporting bias and the challenge it poses for knowledge extraction. We conclude with discussion of approaches to learning commonsense knowledge from text despite this distortion.
While there exist many examples of institutional use of the results of the National Survey of Student Engagement (NSSE), there is a relative paucity of research explicitly linking student outcomes to responses on the survey. A major Doctoral-Extensive institution in the Southeast recently conducted a large-scale implementation of the National Survey of Student Engagement (NSSE). We have linked multiple years of NSSE responses to several student outcomes: freshman retention, GPA, pursuit of graduate education, and employment outcome upon commencement/degree conferral. Our research finds minimal explanatory power in the NSSE benchmarks for these outcomes. A statistically derived model from the individual NSSE items shows greater promise, although there are difficulties in replicating the model for previous student cohorts.
Our goal is to generate reading lists for students that help them optimally learn technical material. Existing retrieval algorithms return items directly relevant to a query but do not return results to help users read about the concepts supporting their query. This is because the dependency structure of concepts that must be understood before reading material pertaining to a given query is never considered. Here we formulate an information-theoretic view of concept dependency and present methods to construct a "concept graph" automatically from a text corpus. We perform the first human evaluation of concept dependency edges (to be published as open data), and the results verify the feasibility of automatic approaches for inferring concepts and their dependency relations. This result can support search capabilities that may be tuned to help users learn a subject rather than retrieve documents based on a single query.
This paper introduces a new framework for data efficient and versatile learning. Specifically: 1) We develop ML-PIP, a general framework for Meta-Learning approximate Probabilistic Inference for Prediction. ML-PIP extends existing probabilistic interpretations of meta-learning to cover a broad class of methods. 2) We introduce VERSA, an instance of the framework employing a flexible and versatile amortization network that takes few-shot learning datasets as inputs, with arbitrary numbers of shots, and outputs a distribution over task-specific parameters in a single forward pass. VERSA substitutes optimization at test time with forward passes through inference networks, amortizing the cost of inference and relieving the need for second derivatives during training. 3) We evaluate VERSA on benchmark datasets where the method sets new state-of-the-art results, handles arbitrary numbers of shots, and for classification, arbitrary numbers of classes at train and test time. The power of the approach is then demonstrated through a challenging few-shot ShapeNet view reconstruction task.
In neural architecture search (NAS), the space of neural network architectures is automatically explored to maximize predictive accuracy for a given task. Despite the success of recent approaches, most existing methods cannot be directly applied to large scale problems because of their prohibitive computational complexity or high memory usage. In this work, we propose a Probabilistic approach to neural ARchitecture SEarCh (PARSEC) that drastically reduces memory requirements while maintaining state-of-the-art computational complexity, making it possible to directly search over more complex architectures and larger datasets. Our approach only requires as much memory as is needed to train a single architecture from our search space. This is due to a memory-efficient sampling procedure wherein we learn a probability distribution over high-performing neural network architectures. Importantly, this framework enables us to transfer the distribution of architectures learnt on smaller problems to larger ones, further reducing the computational cost. We showcase the advantages of our approach in applications to CIFAR-10 and ImageNet, where our approach outperforms methods with double its computational cost and matches the performance of methods with costs that are three orders of magnitude larger.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.