John Foley

Emma Tosch, Kaleigh Clary, John Foley et al.

Evaluation of deep reinforcement learning (RL) is inherently challenging. In particular, learned policies are largely opaque, and hypotheses about the behavior of deep RL agents are difficult to test in black-box environments. Considerable effort has gone into addressing opacity, but almost no effort has been devoted to producing high quality environments for experimental evaluation of agent behavior. We present TOYBOX, a new high-performance, open-source* subset of Atari environments re-designed for the experimental evaluation of deep RL. We show that TOYBOX enables a wide range of experiments and analyses that are impossible in other environments. *https://kdl-umass.github.io/Toybox/

John C. Baez, John Foley

System of systems engineering seeks to analyze, design and deploy collections of systems that together can flexibly address an array of complex tasks. In the Complex Adaptive System Composition and Design Environment program, we developed "network operads" as a tool for designing and tasking systems of systems, and applied them to domains including maritime search and rescue. The network operad formalism offers new ways to handle changing levels of abstraction in system-of-system design and tasking.

Laura Dietz, Bhaskar Mitra, Jeremy Pickens et al.

The vision of HIPstIR is that early stage information retrieval (IR) researchers get together to develop a future for non-mainstream ideas and research agendas in IR. The first iteration of this vision materialized in the form of a three day workshop in Portsmouth, New Hampshire attended by 24 researchers across academia and industry. Attendees pre-submitted one or more topics that they want to pitch at the meeting. Then over the three days during the workshop, we self-organized into groups and worked on six specific proposals of common interest. In this report, we present an overview of the workshop and brief summaries of the six proposals that resulted from the workshop.

Kaleigh Clary, Emma Tosch, John Foley et al.

Reproducibility in reinforcement learning is challenging: uncontrolled stochasticity from many sources, such as the learning algorithm, the learned policy, and the environment itself have led researchers to report the performance of learned agents using aggregate metrics of performance over multiple random seeds for a single environment. Unfortunately, there are still pernicious sources of variability in reinforcement learning agents that make reporting common summary statistics an unsound metric for performance. Our experiments demonstrate the variability of common agents used in the popular OpenAI Baselines repository. We make the case for reporting post-training agent performance as a distribution, rather than a point estimate.

John Foley, Emma Tosch, Kaleigh Clary et al.

It is a widely accepted principle that software without tests has bugs. Testing reinforcement learning agents is especially difficult because of the stochastic nature of both agents and environments, the complexity of state-of-the-art models, and the sequential nature of their predictions. Recently, the Arcade Learning Environment (ALE) has become one of the most widely used benchmark suites for deep learning research, and state-of-the-art Reinforcement Learning (RL) agents have been shown to routinely equal or exceed human performance on many ALE tasks. Since ALE is based on emulation of original Atari games, the environment does not provide semantically meaningful representations of internal game state. This means that ALE has limited utility as an environment for supporting testing or model introspection. We propose ToyBox, a collection of reimplementations of these games that solves this critical problem and enables robust testing of RL agents.

John Foley

The field of information retrieval often works with limited and noisy data in an attempt to classify documents into subjective categories, e.g., relevance, sentiment and controversy. We typically quantify a notion of agreement to understand the difficulty of the labeling task, but when we present final results, we do so using measures that are unaware of agreement or the inherent subjectivity of the task. We propose using user simulation to understand the effect size of this noisy agreement data. By simulating truth and predictions, we can understand the maximum scores a dataset can support: for if a classifier is doing better than a reasonable model of a human, we cannot conclude that it is actually better, but that it may be learning noise present in the dataset. We present a brief case study on controversy detection that concludes that a commonly-used dataset has been exhausted: in order to advance the state-of-the-art, more data must be gathered at the current level of label agreement in order to distinguish between techniques with confidence.

John Foley, Sheikh Muhammad Sarwar, James Allan

Traditional information retrieval treats named entity recognition as a pre-indexing corpus annotation task, allowing entity tags to be indexed and used during search. Named entity taggers themselves are typically trained on thousands or tens of thousands of examples labeled by humans. However, there is a long tail of named entities classes, and for these cases, labeled data may be impossible to find or justify financially. We propose exploring named entity recognition as a search task, where the named entity class of interest is a query, and entities of that class are the relevant "documents". What should that query look like? Can we even perform NER-style labeling with tens of labels? This study presents an exploration of CRF-based NER models with handcrafted features and of how we might transform them into search queries.

Laura Dietz, John Foley

The sequential dependence model (SDM) is a popular retrieval model which is based on the theory of probabilistic graphical models. While it was originally introduced by Metzler and Croft as a Markov Random Field (aka discriminative probabilistic model), in this paper we demonstrate that it is equivalent to a generative probabilistic model. To build an foundation for future retrieval models, this paper details the axiomatic underpinning of the SDM model as discriminative and generative probabilistic model. The only difference arises whether model parameters are estimated in log-space or Multinomial-space. We demonstrate that parameter-estimation with grid-tuning is negatively impacting the generative formulation, an effect that vanishes when parameters are estimated with coordinate-gradient descent. This is concerning, since empirical differences may be falsely attributed to improved models.

Sheikh Muhammad Sarwar, John Foley, James Allan

We address the role of a user in Contextual Named Entity Retrieval (CNER), showing (1) that user identification of important context-bearing terms is superior to automated approaches, and (2) that further gains are possible if the user indicates the relative importance of those terms. CNER is similar in spirit to List Question answering and Entity disambiguation. However, the main focus of CNER is to obtain user feedback for constructing a profile for a class of entities on the fly and use that to retrieve entities from free text. Given a sentence, and an entity selected from that sentence, CNER aims to retrieve sentences that have entities similar to query entity. This paper explores obtaining term relevance feedback and importance weighting from humans in order to improve a CNER system. We report our findings based on the efforts of IR researchers as well as crowdsourced workers.