Search

How do author perceptions match up to the outcomes of the peer-review process and perceptions of others? In a top-tier computer science conference (NeurIPS 2021) with more than 23,000 submitting authors and 9,000 submitted papers, we surveyed the authors on three questions: (i) their predicted probability of acceptance for each of their papers, (ii) their perceived ranking of their own papers based on scientific contribution, and (iii) the change in their perception about their own papers after seeing the reviews. The salient results are: (1) Authors had roughly a three-fold overestimate of the acceptance probability of their papers: The median prediction was 70% for an approximately 25% acceptance rate. (2) Female authors exhibited a marginally higher (statistically significant) miscalibration than male authors; predictions of authors invited to serve as meta-reviewers or reviewers were similarly calibrated, but better than authors who were not invited to review. (3) Authors' relative ranking of scientific contribution of two submissions they made generally agreed with their predicted acceptance probabilities (93% agreement), but there was a notable 7% responses where authors predicted a worse outcome for their better paper. (4) The author-provided rankings disagreed with the peer-review decisions about a third of the time; when co-authors ranked their jointly authored papers, co-authors disagreed at a similar rate—about a third of the time. (5) At least 30% of respondents of both accepted and rejected papers said that their perception of their own paper improved after the review process. The stakeholders in peer review should take these findings into account in setting their expectations from peer review. [ABSTRACT FROM AUTHOR]

Peer review is the backbone of academia and humans constitute a cornerstone of this process, being responsible for reviewing submissions and making the final acceptance/rejection decisions. Given that human decision-making is known to be susceptible to various cognitive biases, it is important to understand which (if any) biases are present in the peer-review process, and design the pipeline such that the impact of these biases is minimized. In this work, we focus on the dynamics of discussions between reviewers and investigate the presence of herding behaviour therein. Specifically, we aim to understand whether reviewers and discussion chairs get disproportionately influenced by the first argument presented in the discussion when (in case of reviewers) they form an independent opinion about the paper before discussing it with others. In conjunction with the review process of a large, top tier machine learning conference, we design and execute a randomized controlled trial that involves 1,544 papers and 2,797 reviewers with the goal of testing for the conditional causal effect of the discussion initiator's opinion on the outcome of a paper. Our experiment reveals no evidence of herding in peer-review discussions. This observation is in contrast with past work that has documented an undue influence of the first piece of information on the final decision (e.g., anchoring effect) and analyzed herding behaviour in other applications (e.g., financial markets). Regarding policy implications, the absence of the herding effect suggests that the current status quo of the absence of a unified policy towards discussion initiation does not result in an increased arbitrariness of the resulting decisions. [ABSTRACT FROM AUTHOR]

Citations play an important role in researchers' careers as a key factor in evaluation of scientific impact. Many anecdotes advice authors to exploit this fact and cite prospective reviewers to try obtaining a more positive evaluation for their submission. In this work, we investigate if such a citation bias actually exists: Does the citation of a reviewer's own work in a submission cause them to be positively biased towards the submission? In conjunction with the review process of two flagship conferences in machine learning and algorithmic economics, we execute an observational study to test for citation bias in peer review. In our analysis, we carefully account for various confounding factors such as paper quality and reviewer expertise, and apply different modeling techniques to alleviate concerns regarding the model mismatch. Overall, our analysis involves 1,314 papers and 1,717 reviewers and detects citation bias in both venues we consider. In terms of the effect size, by citing a reviewer's work, a submission has a non-trivial chance of getting a higher score from the reviewer: an expected increase in the score is approximately 0.23 on a 5-point Likert item. For reference, a one-point increase of a score by a single reviewer improves the position of a submission by 11% on average. [ABSTRACT FROM AUTHOR]

There is widespread debate on whether to anonymize author identities in peer review. The key argument for anonymization is to mitigate bias, whereas arguments against anonymization posit various uses of author identities in the review process. The Innovations in Theoretical Computer Science (ITCS) 2023 conference adopted a middle ground by initially anonymizing the author identities from reviewers, revealing them after the reviewer had submitted their initial reviews, and allowing the reviewer to change their review subsequently. We present an analysis of the reviews pertaining to the identification and use of author identities. Our key findings are: (I) A majority of reviewers self-report not knowing and being unable to guess the authors' identities for the papers they were reviewing. (II) After the initial submission of reviews, 7.1% of reviews changed their overall merit score and 3.8% changed their self-reported reviewer expertise. (III) There is a very weak and statistically insignificant correlation of the rank of authors' affiliations with the change in overall merit; there is a weak but statistically significant correlation with respect to change in reviewer expertise. We also conducted an anonymous survey to obtain opinions from reviewers and authors. The main findings from the 200 survey responses are: (i) A vast majority of participants favor anonymizing author identities in some form. (ii) The "middle-ground" initiative of ITCS 2023 was appreciated. (iii) Detecting conflicts of interest is a challenge that needs to be addressed if author identities are anonymized. Overall, these findings support anonymization of author identities in some form (e.g., as was done in ITCS 2023), as long as there is a robust and efficient way to check conflicts of interest. [ABSTRACT FROM AUTHOR]

The task of aggregating and denoising crowd-labeled data has gained increased significance with the advent of crowdsourcing platforms and massive datasets. We propose a permutation-based model for crowd labeled data that is a significant generalization of the classical Dawid-Skene model, and introduce a new error metric by which to compare different estimators. We derive global minimax rates for the permutation-based model that are sharp up to logarithmic factors, and match the minimax lower bounds derived under the simpler Dawid-Skene model. We then design two computationally-efficient estimators: the WAN estimator for the setting where the ordering of workers in terms of their abilities is approximately known, and the OBI- WAN estimator where that is not known. For each of these estimators, we provide non-asymptotic bounds on their performance. We conduct synthetic simulations and experiments on real-world crowdsourcing data, and the experimental results corroborate our theoretical findings. [ABSTRACT FROM AUTHOR]

We study methods for aggregating pairwise comparison data among a collection of $n$ items with the goal of estimating the outcome probabilities for future comparisons. Working within a flexible model that only imposes a form of strong stochastic transitivity, we introduce an “adaptivity index” which compares the risk of our estimator to that of an oracle, over appropriate sub-models, where the oracle knows the specific sub-model in the ground truth. In addition to measuring the usual worst-case risk of an estimator, this adaptivity index also captures the extent to which the estimator adapts to instance-specific difficulty relative to an oracle estimator. First, we propose a three-step estimator termed count-randomize-least squares, and show that it has adaptivity index upper bounded by $\sqrt {n}$ up to logarithmic factors. We then show that conditional on the planted clique hypothesis, no computationally efficient estimator can achieve an adaptivity index smaller than $\sqrt {n}$. Second, we show that a regularized least squares estimator can achieve a poly-logarithmic adaptivity index, thereby demonstrating a $\sqrt {n}$ -gap between optimal and computationally achievable adaptivity. Finally, we prove that the standard least squares estimator, which is known to be optimally adaptive in several closely related problems, fails to adapt in the context of estimating pairwise probabilities. [ABSTRACT FROM AUTHOR]

In the above article , we introduced two scheduling policies and analyzed their average job latencies. With an implicit assumption that the scheduling policies provide sample-path bounds by construction, we claimed that their average job latencies serve as upper and lower bounds on that of a centralized MDS queue. In this note, we present recently discovered counterexamples, disproving the assumption. We replace the assumption with a conjecture that the average latency bounds still hold. We also provide an erratum to the original article to correct any confusing or misleading statements. [ABSTRACT FROM AUTHOR]

We consider data in the form of pairwise comparisons of n items, with the goal of identifying the top k items for some value of k < n, or alternatively, recovering a ranking of all the items. We analyze the Borda counting algorithm that ranks the items in order of the number of pairwise comparisons won, and show it has three attractive features: (a) it is an optimal method achieving the information-theoretic limits up to constant factors; (b) it is robust in that its optimality holds without imposing conditions on the underlying matrix of pairwise-comparison probabilities, in contrast to some prior work that applies only to the BTL parametric model; and (c) its computational efficiency leads to speed-ups of several orders of magnitude. We address the problem of exact recovery, and for the top-k recovery problem we also extend our results to obtain sharp guarantees for approximate recovery under the Hamming distortion metric, and more generally, to any arbitrary error requirement that satisfies a simple and natural monotonicity condition. In doing so, we introduce a general framework that allows us to treat a variety of problems in the literature in an unified manner. [ABSTRACT FROM AUTHOR]

Repair operations in erasure-coded distributed storage systems involve a lot of data movement. This can potentially expose data to malicious acts of passive eavesdroppers or active adversaries, putting security of the system at risk. This paper presents coding schemes and repair algorithms that ensure security of the data in the presence of passive eavesdroppers and active adversaries while maintaining high availability, reliability, and resource efficiency in the system. The proposed codes are optimal in that they meet previously proposed lower bounds on storage and network-bandwidth requirements for a wide range of system parameters. The results thus establish the secure storage capacity of such systems. The proposed codes are based on an optimal class of codes called product-matrix codes. The constructions presented for security from active adversaries provide an additional appealing feature of “on-demand security,” where the desired level of security can be chosen separately for each instance of repair, and the proposed algorithms remain optimal simultaneously for all possible security levels. This paper also provides necessary and sufficient conditions governing the transformation of any (non-secure) code into one providing on-demand security. [ABSTRACT FROM PUBLISHER]

We consider a problem of prediction based on opinions elicited from heterogeneous rational agents with private information. Making an accurate prediction with a minimal cost requires a joint design of the incentive mechanism and the prediction algorithm. To elicit heterogeneous agents' private information and incentivize agents with different capabilities to act in the principal's best interest,we design an optimal joint incentive mechanism and prediction algorithm called COPE (COst and Prediction Elicitation), the analysis of which offers several valuable engineering insights. First, when the costs incurred by the agents are linear in the exerted effort, COPE corresponds to a "crowdcontending" mechanism, where the principal only employs the agent with the highest capability. Second, when the costs are quadratic, COPE corresponds to a "crowdsourcing" mechanism that employs multiple agents with different capabilities at the same time. Numerical simulations show that COPE improves the principal's profit (The improvement is 5%-30% in our simulations), comparing to those mechanisms that assume all agents have equal capabilities. [ABSTRACT FROM AUTHOR]

Erasure codes are being extensively deployed in distributed storage systems instead of replication to achieve fault tolerance in a storage efficient manner. While traditional erasure codes are storage efficient, they can result in a significant increase in the amount of data access and downloaded during rebuilding of failed or otherwise unavailable nodes. In this paper, we present a new framework, which we call piggybacking, for constructing distributed storage codes that are efficient in the amount of data read and downloaded during rebuilding, while meeting requirements arising out of system considerations in data centers—maximum-distance-separability (MDS), high-rate, and a small number of so-called substripes. Under this setting, to the best of our knowledge, piggyback codes achieve the minimum average amount of data access and downloaded during rebuilding among all existing explicit solutions. The piggybacking framework also offers a rich design space for constructing codes for a variety of other settings. In particular, we construct codes that require minimum amount of data access and downloaded for rebuilding among all existing solutions for: 1) binary MDS array codes with more than two parities and 2) MDS codes with the smallest locality during rebuilding. In addition, we show how piggybacking can be employed to enable efficient repair of parity nodes in codes that address the rebuilding of only systematic nodes. The basic idea behind the piggybacking framework is to take multiple instances of existing codes and add carefully designed functions of the data from one instance to the others. This framework provides 25% to 50% savings in the average amount of data access and downloaded during rebuilding depending on the choice of the code parameters. [ABSTRACT FROM AUTHOR]

In order to scale economically, data centers are increasingly evolving their data storage methods from simple data replication to more powerful erasure codes, which provide the same level of reliability as replication, but at a significantly lower storage cost. In particular, it is well known that maximum-distance-separable (MDS) codes, such as Reed–Solomon codes, can achieve a target reliability with the maximum storage efficiency. While the use of codes for providing improved reliability in archival storage systems, where data is less frequently accessed (or so-called “cold data”), is well understood, the role of codes in storing more frequently accessed and active “hot data”, where latency is the key metric, is less clear. In this paper, we study data storage systems based on MDS codes through the lens of queueing theory, and term the queueing system arising under codes as an “MDS queue.” We provide lower and upper bounds on the average job latency for both centralized and decentralized versions of MDS queues. We also provide extensive simulations to corroborate our analysis as well as obtain additional insights. [ABSTRACT FROM PUBLISHER]

There are various parametric models for analyzing pairwise comparison data, including the Bradley–Terry–Luce (BTL) and Thurstone models, but their reliance on strong parametric assumptions is limiting. In this paper, we study a flexible model for pairwise comparisons, under which the probabilities of outcomes are required only to satisfy a natural form of stochastic transitivity. This class includes parametric models, including the BTL and Thurstone models as special cases, but is considerably more general. We provide various examples of models in this broader stochastically transitive class for which classical parametric models provide poor fits. Despite this greater flexibility, we show that the matrix of probabilities can be estimated at the same rate as in standard parametric models up to logarithmic terms. On the other hand, unlike in the BTL and Thurstone models, computing the minimax-optimal estimator in the stochastically transitive model is non-trivial, and we explore various computationally tractable alternatives. We show that a simple singular value thresholding algorithm is statistically consistent but does not achieve the minimax rate. We then propose and study algorithms that achieve the minimax rate over interesting sub-classes of the full stochastically transitive class. We complement our theoretical results with thorough numerical simulations. [ABSTRACT FROM AUTHOR]

Many systems possess the flexibility to serve requests in more than one way, such as distributed storage systems that store multiple copies of the data. In such systems, the latency of serving the requests may potentially be reduced by sending redundant requests: a request may be sent to more servers than needed and deemed served when the requisite number of servers complete service. Such a mechanism trades off the possibility of faster execution of the request with the increase in the load on the system. Several recent works empirically evaluate the latency performance of redundant requests in diverse settings. In this paper, we perform an analytical study of the latency performance of redundant requests, with the primary goals of characterizing under what scenarios sending redundant requests will help (and under what scenarios it will not), and of designing optimal redundant-requesting policies. We show that when service times are i.i.d. memoryless or “heavier,” and when the additional copies of already-completed jobs can be removed instantly, maximally scheduling redundant requests achieves the optimal average latency. On the other hand, when service times are i.i.d. “lighter” or when service times are memoryless and removal of jobs is not instantaneous, then not having any redundancy in the requests is optimal under high loads. Our results are applicable to arbitrary arrival processes. [ABSTRACT FROM AUTHOR]

Shamir's (n,k) threshold secret sharing is an important component of several cryptographic protocols, such as those for secure multiparty-computation and key management. These protocols typically assume the presence of direct communication links from the dealer to all participants, in which case the dealer can directly pass the shares of the secret to each participant. In this paper, we consider the problem of secret sharing when the dealer does not have direct communication links to all the participants, and instead, the dealer and the participants form a general network. Existing methods are based on separate secure message transmissions from the dealer to each participant, requiring considerable coordination and communication in the network. We present a distributed algorithm for disseminating shares over a network, which we call the “SNEAK” algorithm, requiring each node to know only the identities of its one-hop neighbors. While SNEAK imposes a stronger condition on the network by requiring the dealer to be what we call k-propagating rather than k-connected as required by the existing solutions, we show that in addition to being distributed, it achieves significant reduction in the amount of communication and the randomness required. We also derive information-theoretic lower bounds on the amount of communication for secret sharing over networks, which may be of independent interest. [ABSTRACT FROM PUBLISHER]

We consider the problem of efficient recovery of the data stored in any individual node of a distributed storage system, from the rest of the nodes. Applications include handling failures and degraded reads. We measure efficiency in terms of the amount of data-read and the download required. To minimize the download, we focus on the minimum bandwidth setting of the `regenerating codes' model for distributed storage. Under this model, the system has a total of n nodes, and the data stored in any node must be (efficiently) recoverable from any d of the other (n-1) nodes. Lower bounds on the two metrics under this model were derived previously; it has also been shown that these bounds are achievable for the amount of data-read and download when d=n-1, and for the amount of download alone when d\neq n-1. In this paper, we complete the picture by proving the converse result, that when d\neq n-1, these lower bounds are strictly loose with respect to the amount of read required. The proof is information-theoretic, and hence applies to non-linear codes as well. We also show that under two (practical) relaxations of the problem setting, these lower bounds can be met for both read and download simultaneously. [ABSTRACT FROM PUBLISHER]

Regenerating codes are a class of recently developed codes for distributed storage that, like Reed-Solomon codes, permit data recovery from any arbitrary k of n nodes. However regenerating codes possess in addition, the ability to repair a failed node by connecting to any arbitrary d nodes and downloading an amount of data that is typically far less than the size of the data file. This amount of download is termed the repair bandwidth. Minimum storage regenerating (MSR) codes are a subclass of regenerating codes that require the least amount of network storage; every such code is a maximum distance separable (MDS) code. Further, when a replacement node stores data identical to that in the failed node, the repair is termed as exact. [ABSTRACT FROM AUTHOR]

Regenerating codes are a class of recently developed codes for distributed storage that, like Reed-Solomon codes, permit data recovery from any subset of k nodes within the n-node network. However, regenerating codes possess in addition, the ability to repair a failed node by connecting to an arbitrary subset of d nodes. It has been shown that for the case of functional repair, there is a tradeoff between the amount of data stored per node and the bandwidth required to repair a failed node. A special case of functional repair is exact repair where the replacement node is required to store data identical to that in the failed node. Exact repair is of interest as it greatly simplifies system implementation. The first result of this paper is an explicit, exact-repair code for the point on the storage-bandwidth tradeoff corresponding to the minimum possible repair bandwidth, for the case when d=n-1. This code has a particularly simple graphical description, and most interestingly has the ability to carry out exact repair without any need to perform arithmetic operations. We term this ability of the code to perform repair through mere transfer of data as repair by transfer. The second result of this paper shows that the interior points on the storage-bandwidth tradeoff cannot be achieved under exact repair, thus pointing to the existence of a separate tradeoff under exact repair. Specifically, we identify a set of scenarios which we term as “helper node pooling,” and show that it is the necessity to satisfy such scenarios that overconstrains the system. [ABSTRACT FROM AUTHOR]

Regenerating codes are a class of distributed storage codes that allow for efficient repair of failed nodes, as compared to traditional erasure codes. An [n, k, d] regenerating code permits the data to be recovered by connecting to any k of the n nodes in the network, while requiring that a failed node be repaired by connecting to any d nodes. The amount of data downloaded for repair is typically much smaller than the size of the source data. Previous constructions of exact-regenerating codes have been confined to the case n=d+1. In this paper, we present optimal, explicit constructions of (a) Minimum Bandwidth Regenerating (MBR) codes for all values of [n, k, d] and (b) Minimum Storage Regenerating (MSR) codes for all [n, k, d\geq 2k-2], using a new product-matrix framework. The product-matrix framework is also shown to significantly simplify system operation. To the best of our knowledge, these are the first constructions of exact-regenerating codes that allow the number n of nodes in the network, to be chosen independent of the other parameters. The paper also contains a simpler description, in the product-matrix framework, of a previously constructed MSR code with [n=d+1, k, d\geq 2k-1]. [ABSTRACT FROM PUBLISHER]

A number of applications require two-sample testing on ranked preference data. For instance, in crowdsourcing, there is a long-standing question of whether pairwise-comparison data provided by people is distributed identically to ratings-converted-to-comparisons. Other applications include sports data analysis and peer grading. In this paper, we design twosample tests for pairwise-comparison data and ranking data. For our two-sample test for pairwise-comparison data, we establish an upper bound on the sample complexity required to correctly test whether the distributions of the two sets of samples are identical. Our test requires essentially no assumptions on the distributions. We then prove complementary lower bounds showing that our results are tight (in the minimax sense) up to constant factors. We investigate the role of modeling assumptions by proving lower bounds for a range of pairwise-comparison models (WST, MST, SST, parameter-based such as BTL and Thurstone). We also provide tests and associated sample complexity bounds for partial (or total) ranking data. Furthermore, we empirically evaluate our results via extensive simulations as well as three real-world data sets consisting of pairwise-comparisons and rankings. By applying our two-sample test on real-world pairwise-comparison data, we conclude that ratings and rankings provided by people are indeed distributed differently. [ABSTRACT FROM AUTHOR]