arXiv subjects

Robert West

Publications and source records attributed to Robert West.

At least 163 records · Page 9Linked to original sources

arXiv2020

Global gender differences in Wikipedia readership

Wikipedia represents the largest and most popular source of encyclopedic knowledge in the world today, aiming to provide equal access to information worldwide. From a global online survey of 65,031 readers of Wikipedia and their corresponding reading logs, we present novel evidence of gender differences in Wikipedia readership and how they manifest in records of user behavior. More specifically we report that (1) women are underrepresented among readers of Wikipedia, (2) women view fewer pages per reading session than men do, (3) men and women visit Wikipedia for similar reasons, and (4) men and women exhibit specific topical preferences. Our findings lay the foundation for identifying pathways toward knowledge equity in the usage of online encyclopedic knowledge.

cs.CY↗

arXiv2020

On the Limitations of Cross-lingual Encoders as Exposed by Reference-Free Machine Translation Evaluation

Evaluation of cross-lingual encoders is usually performed either via zero-shot cross-lingual transfer in supervised downstream tasks or via unsupervised cross-lingual textual similarity. In this paper, we concern ourselves with reference-free machine translation (MT) evaluation where we directly compare source texts to (sometimes low-quality) system translations, which represents a natural adversarial setup for multilingual encoders. Reference-free evaluation holds the promise of web-scale comparison of MT systems. We systematically investigate a range of metrics based on state-of-the-art cross-lingual semantic representations obtained with pretrained M-BERT and LASER. We find that they perform poorly as semantic encoders for reference-free MT evaluation and identify their two key limitations, namely, (a) a semantic mismatch between representations of mutual translations and, more prominently, (b) the inability to punish "translationese", i.e., low-quality literal translations. We propose two partial remedies: (1) post-hoc re-alignment of the vector spaces and (2) coupling of semantic-similarity based metrics with target-side language modeling. In segment-level MT evaluation, our best metric surpasses reference-based BLEU by 5.7 correlation points.

cs.CL↗

arXiv2020

Robust Cross-lingual Embeddings from Parallel Sentences

Recent advances in cross-lingual word embeddings have primarily relied on mapping-based methods, which project pretrained word embeddings from different languages into a shared space through a linear transformation. However, these approaches assume word embedding spaces are isomorphic between different languages, which has been shown not to hold in practice (Søgaard et al., 2018), and fundamentally limits their performance. This motivates investigating joint learning methods which can overcome this impediment, by simultaneously learning embeddings across languages via a cross-lingual term in the training objective. We propose a bilingual extension of the CBOW method which leverages sentence-aligned corpora to obtain robust cross-lingual word and sentence representations. Our approach significantly improves cross-lingual sentence retrieval performance over all other approaches while maintaining parity with the current state-of-the-art methods on word-translation. It also achieves parity with a deep RNN method on a zero-shot cross-lingual document classification task, requiring far fewer computational resources for training and inference. As an additional advantage, our bilingual method leads to a much more pronounced improvement in the the quality of monolingual word vectors compared to other competing methods.

cs.CL↗

arXiv2020

WikiHist.html: English Wikipedia's Full Revision History in HTML Format

Wikipedia is written in the wikitext markup language. When serving content, the MediaWiki software that powers Wikipedia parses wikitext to HTML, thereby inserting additional content by expanding macros (templates and mod-ules). Hence, researchers who intend to analyze Wikipediaas seen by its readers should work with HTML, rather than wikitext. Since Wikipedia's revision history is publicly available exclusively in wikitext format, researchers have had to produce HTML themselves, typically by using Wikipedia's REST API for ad-hoc wikitext-to-HTML parsing. This approach, however, (1) does not scale to very large amounts ofdata and (2) does not correctly expand macros in historical article revisions. We solve these problems by developing a parallelized architecture for parsing massive amounts of wikitext using local instances of MediaWiki, enhanced with the capacity of correct historical macro expansion. By deploying our system, we produce and release WikiHist.html, English Wikipedia's full revision history in HTML format. We highlight the advantages of WikiHist.html over raw wikitext in an empirical analysis of Wikipedia's hyperlinks, showing that over half of the wiki links present in HTML are missing from raw wikitext and that the missing links are important for user navigation.

cs.CY↗

arXiv2020

Fixation properties of rock-paper-scissors games in fluctuating populations

Rock-paper-scissors games metaphorically model cyclic dominance in ecology and microbiology. In a static environment, these models are characterized by fixation probabilities obeying two different "laws" in large and small well-mixed populations. Here, we investigate the evolution of these three-species models subject to a randomly switching carrying capacity modeling the endless change between states of resources scarcity and abundance. Focusing mainly on the zero-sum rock-paper-scissors game, equivalent to the cyclic Lotka-Volterra model, we study how the ${\it coupling}$ of demographic and environmental noise influences the fixation properties. More specifically, we investigate which species is the most likely to prevail in a population of fluctuating size and how the outcome depends on the environmental variability. We show that demographic noise coupled with environmental randomness "levels the field" of cyclic competition by balancing the effect of selection. In particular, we show that fast switching effectively reduces the selection intensity proportionally to the variance of the carrying capacity. We determine the conditions under which new fixation scenarios arise, where the most likely species to prevail changes with the rate of switching and the variance of the carrying capacity. Random switching has a limited effect on the mean fixation time that scales linearly with the average population size. Hence, environmental randomness makes the cyclic competition more egalitarian, but does not prolong the species coexistence. We also show how the fixation probabilities of close-to-zero-sum rock-paper-scissors games can be obtained from those of the zero-sum model by rescaling the selection intensity.

q-bio.PE↗

arXiv2020

Behavior Cloning in OpenAI using Case Based Reasoning

Learning from Observation (LfO), also known as Behavioral Cloning, is an approach for building software agents by recording the behavior of an expert (human or artificial) and using the recorded data to generate the required behavior. jLOAF is a platform that uses Case-Based Reasoning to achieve LfO. In this paper we interface jLOAF with the popular OpenAI Gym environment. Our experimental results show how our approach can be used to provide a baseline for comparison in this domain, as well as identify the strengths and weaknesses when dealing with environmental complexity.

cs.AI↗

arXiv2020

Learning High Order Feature Interactions with Fine Control Kernels

We provide a methodology for learning sparse statistical models that use as features all possible multiplicative interactions among an underlying atomic set of features. While the resulting optimization problems are exponentially sized, our methodology leads to algorithms that can often solve these problems exactly or provide approximate solutions based on combining highly correlated features. We also introduce an algorithmic paradigm, the Fine Control Kernel framework, so named because it is based on Fenchel Duality and is reminiscent of kernel methods. Its theory is tailored to large sparse learning problems, and it leads to efficient feature screening rules for interactions. These rules are inspired by the Apriori algorithm for market basket analysis -- which also falls under the purview of Fine Control Kernels, and can be applied to a plurality of learning problems including the Lasso and sparse matrix estimation. Experiments on biomedical datasets demonstrate the efficacy of our methodology in deriving algorithms that efficiently produce interactions models which achieve state-of-the-art accuracy and are interpretable.

cs.LG↗

arXiv2020

Quantifying Engagement with Citations on Wikipedia

Wikipedia, the free online encyclopedia that anyone can edit, is one of the most visited sites on the Web and a common source of information for many users. As an encyclopedia, Wikipedia is not a source of original information, but was conceived as a gateway to secondary sources: according to Wikipedia's guidelines, facts must be backed up by reliable sources that reflect the full spectrum of views on the topic. Although citations lie at the very heart of Wikipedia, little is known about how users interact with them. To close this gap, we built client-side instrumentation for logging all interactions with links leading from English Wikipedia articles to cited references during one month, and conducted the first analysis of readers' interaction with citations on Wikipedia. We find that overall engagement with citations is low: about one in 300 page views results in a reference click (0.29% overall; 0.56% on desktop; 0.13% on mobile). Matched observational studies of the factors associated with reference clicking reveal that clicks occur more frequently on shorter pages and on pages of lower quality, suggesting that references are consulted more commonly when Wikipedia itself does not contain the information sought by the user. Moreover, we observe that recent content, open access sources and references about life events (births, deaths, marriages, etc) are particularly popular. Taken together, our findings open the door to a deeper understanding of Wikipedia's role in a global information economy where reliability is ever less certain, and source attribution ever more vital.

cs.CY↗

arXiv2019

Deep Learning for Prostate Pathology

The current study detects different morphologies related to prostate pathology using deep learning models; these models were evaluated on 2,121 hematoxylin and eosin (H&E) stain histology images captured using bright field microscopy, which spanned a variety of image qualities, origins (whole slide, tissue micro array, whole mount, Internet), scanning machines, timestamps, H&E staining protocols, and institutions. For case usage, these models were applied for the annotation tasks in clinician-oriented pathology reports for prostatectomy specimens. The true positive rate (TPR) for slides with prostate cancer was 99.7% by a false positive rate of 0.785%. The F1-scores of Gleason patterns reported in pathology reports ranged from 0.795 to 1.0 at the case level. TPR was 93.6% for the cribriform morphology and 72.6% for the ductal morphology. The correlation between the ground truth and the prediction for the relative tumor volume was 0.987 n. Our models cover the major components of prostate pathology and successfully accomplish the annotation tasks.

q-bio.TO↗

arXiv2019

Registration of pre-surgical MRI and whole-mount histopathology images in prostate cancer patients with radical prostatectomy via RAPSODI

Magnetic resonance imaging (MRI) has great potential to improve prostate cancer diagnosis. It can spare men with a normal exam from undergoing invasive biopsy while making biopsies more accurate in men with lesions suspicious for cancer. Yet, the subtle differences between cancer and confounding conditions, render the interpretation of MRI challenging. The tissue collected from patients that undergo pre-surgical MRI and radical prostatectomy provides a unique opportunity to correlate histopathology images of the entire prostate with MRI in order to accurately map the extent of prostate cancer onto MRI. Here, we introduce the RAPSODI (framework for the registration of radiology and pathology images. RAPSODI relies on a three-step procedure that first reconstructs in 3D the resected tissue using the serial whole-mount histopathology slices, then registers corresponding histopathology and MRI slices, and finally maps the cancer outlines from the histopathology slices onto MRI. We tested RAPSODI in a phantom study where we simulated various conditions, e.g., tissue specimen rotation upon mounting on glass slides, tissue shrinkage during fixation, or imperfect slice-to-slice correspondences between histology and MRI. Our experiments showed that RAPSODI can reliably correct for rotations within $\pm15^{\circ}$ and shrinkage up to 10%. We also evaluated RAPSODI in 89 patients from two institutions that underwent radical prostatectomy, yielding 543 histopathology slices that were registered to corresponding T2 weighted MRI slices. We found a Dice coefficient of 0.98$ \pm $0.01 for the prostate, prostate boundary Hausdorff distance of 1.71$ \pm $0.48 mm, a urethra deviation of 2.91$ \pm $1.25 mm, and a landmark deviation of 2.88$ \pm $0.70 mm between registered histopathology images and MRI. Our robust framework successfully mapped the extent of disease from histopathology slices onto MRI.

eess.IV↗

arXiv2019

Causal Effects of Brevity on Style and Success in Social Media

In online communities, where billions of people strive to propagate their messages, understanding how wording affects success is of primary importance. In this work, we are interested in one particularly salient aspect of wording: brevity. What is the causal effect of brevity on message success? What are the linguistic traits of brevity? When is brevity beneficial, and when is it not? Whereas most prior work has studied the effect of wording on style and success in observational setups, we conduct a controlled experiment, in which crowd workers shorten social media posts to prescribed target lengths and other crowd workers subsequently rate the original and shortened versions. This allows us to isolate the causal effect of brevity on the success of a message. We find that concise messages are on average more successful than the original messages up to a length reduction of 30-40%. The optimal reduction is on average between 10% and 20%. The observed effect is robust across different subpopulations of raters and is the strongest for raters who visit social media on a daily basis. Finally, we discover unique linguistic and content traits of brevity and correlate them with the measured probability of success in order to distinguish effective from ineffective shortening strategies. Overall, our findings are important for developing a better understanding of the effect of brevity on the success of messages in online social media.

cs.SI↗

arXiv2019

Reverse-Engineering Satire, or "Paper on Computational Humor Accepted Despite Making Serious Advances"

Humor is an essential human trait. Efforts to understand humor have called out links between humor and the foundations of cognition, as well as the importance of humor in social engagement. As such, it is a promising and important subject of study, with relevance for artificial intelligence and human-computer interaction. Previous computational work on humor has mostly operated at a coarse level of granularity, e.g., predicting whether an entire sentence, paragraph, document, etc., is humorous. As a step toward deep understanding of humor, we seek fine-grained models of attributes that make a given text humorous. Starting from the observation that satirical news headlines tend to resemble serious news headlines, we build and analyze a corpus of satirical headlines paired with nearly identical but serious headlines. The corpus is constructed via Unfun.me, an online game that incentivizes players to make minimal edits to satirical headlines with the goal of making other players believe the results are serious headlines. The edit operations used to successfully remove humor pinpoint the words and concepts that play a key role in making the original, satirical headline funny. Our analysis reveals that the humor tends to reside toward the end of headlines, and primarily in noun phrases, and that most satirical headlines follow a certain logical pattern, which we term false analogy. Overall, this paper deepens our understanding of the syntactic and semantic structure of satirical news headlines and provides insights for building humor-producing systems.

cs.AI↗

arXiv2019

Message Distortion in Information Cascades

Information diffusion is usually modeled as a process in which immutable pieces of information propagate over a network. In reality, however, messages are not immutable, but may be morphed with every step, potentially entailing large cumulative distortions. This process may lead to misinformation even in the absence of malevolent actors, and understanding it is crucial for modeling and improving online information systems. Here, we perform a controlled, crowdsourced experiment in which we simulate the propagation of information from medical research papers. Starting from the original abstracts, crowd workers iteratively shorten previously produced summaries to increasingly smaller lengths. We also collect control summaries where the original abstract is compressed directly to the final target length. Comparing cascades to controls allows us to separate the effect of the length constraint from that of accumulated distortion. Via careful manual coding, we annotate lexical and semantic units in the medical abstracts and track them along cascades. We find that iterative summarization has a negative impact due to the accumulation of error, but that high-quality intermediate summaries result in less distorted messages than in the control case. Different types of information behave differently; in particular, the conclusion of a medical abstract (i.e., its key message) is distorted most. Finally, we compare abstractive with extractive summaries, finding that the latter are less prone to semantic distortion. Overall, this work is a first step in studying information cascades without the assumption that disseminated content is immutable, with implications on our understanding of the role of word-of-mouth effects on the misreporting of science.

cs.SI↗

arXiv2019

Eliciting New Wikipedia Users' Interests via Automatically Mined Questionnaires: For a Warm Welcome, Not a Cold Start

Every day, thousands of users sign up as new Wikipedia contributors. Once joined, these users have to decide which articles to contribute to, which users to seek out and learn from or collaborate with, etc. Any such task is a hard and potentially frustrating one given the sheer size of Wikipedia. Supporting newcomers in their first steps by recommending articles they would enjoy editing or editors they would enjoy collaborating with is thus a promising route toward converting them into long-term contributors. Standard recommender systems, however, rely on users' histories of previous interactions with the platform. As such, these systems cannot make high-quality recommendations to newcomers without any previous interactions -- the so-called cold-start problem. The present paper addresses the cold-start problem on Wikipedia by developing a method for automatically building short questionnaires that, when completed by a newly registered Wikipedia user, can be used for a variety of purposes, including article recommendations that can help new editors get started. Our questionnaires are constructed based on the text of Wikipedia articles as well as the history of contributions by the already onboarded Wikipedia editors. We assess the quality of our questionnaire-based recommendations in an offline evaluation using historical data, as well as an online evaluation with hundreds of real Wikipedia newcomers, concluding that our method provides cohesive, human-readable questions that perform well against several baselines. By addressing the cold-start problem, this work can help with the sustainable growth and maintenance of Wikipedia's diverse editor community.

cs.IR↗

arXiv2019

Crosslingual Document Embedding as Reduced-Rank Ridge Regression

There has recently been much interest in extending vector-based word representations to multiple languages, such that words can be compared across languages. In this paper, we shift the focus from words to documents and introduce a method for embedding documents written in any language into a single, language-independent vector space. For training, our approach leverages a multilingual corpus where the same concept is covered in multiple languages (but not necessarily via exact translations), such as Wikipedia. Our method, Cr5 (Crosslingual reduced-rank ridge regression), starts by training a ridge-regression-based classifier that uses language-specific bag-of-word features in order to predict the concept that a given document is about. We show that, when constraining the learned weight matrix to be of low rank, it can be factored to obtain the desired mappings from language-specific bags-of-words to language-independent embeddings. As opposed to most prior methods, which use pretrained monolingual word vectors, postprocess them to make them crosslingual, and finally average word vectors to obtain document vectors, Cr5 is trained end-to-end and is thus natively crosslingual as well as document-level. Moreover, since our algorithm uses the singular value decomposition as its core operation, it is highly scalable. Experiments show that our method achieves state-of-the-art performance on a crosslingual document retrieval task. Finally, although not trained for embedding sentences and words, it also achieves competitive performance on crosslingual sentence and word retrieval tasks.

cs.CL↗

arXiv2019

Hot Streaks on Social Media

Measuring the impact and success of human performance is common in various disciplines, including art, science, and sports. Quantifying impact also plays a key role on social media, where impact is usually defined as the reach of a user's content as captured by metrics such as the number of views, likes, retweets, or shares. In this paper, we study entire careers of Twitter users to understand properties of impact. We show that user impact tends to have certain characteristics: First, impact is clustered in time, such that the most impactful tweets of a user appear close to each other. Second, users commonly have 'hot streaks' of impact, i.e., extended periods of high-impact tweets. Third, impact tends to gradually build up before, and fall off after, a user's most impactful tweet. We attempt to explain these characteristics using various properties measured on social media, including the user's network, content, activity, and experience, and find that changes in impact are associated with significant changes in these properties. Our findings open interesting avenues for future research on virality and influence on social media.

cs.SI↗

arXiv2019

Expanding the Text Classification Toolbox with Cross-Lingual Embeddings

Most work in text classification and Natural Language Processing (NLP) focuses on English or a handful of other languages that have text corpora of hundreds of millions of words. This is creating a new version of the digital divide: the artificial intelligence (AI) divide. Transfer-based approaches, such as Cross-Lingual Text Classification (CLTC) - the task of categorizing texts written in different languages into a common taxonomy, are a promising solution to the emerging AI divide. Recent work on CLTC has focused on demonstrating the benefits of using bilingual word embeddings as features, relegating the CLTC problem to a mere benchmark based on a simple averaged perceptron. In this paper, we explore more extensively and systematically two flavors of the CLTC problem: news topic classification and textual churn intent detection (TCID) in social media. In particular, we test the hypothesis that embeddings with context are more effective, by multi-tasking the learning of multilingual word embeddings and text classification; we explore neural architectures for CLTC; and we move from bi- to multi-lingual word embeddings. For all architectures, types of word embeddings and datasets, we notice a consistent gain trend in favor of multilingual joint training, especially for low-resourced languages.

cs.CL↗

arXiv2019

The Super-Earth Opportunity - Search for Habitable Exoplanets in the 2020s

The recent discovery of a staggering diversity of planets beyond the Solar System has brought with it a greatly expanded search space for habitable worlds. The Kepler exoplanet survey has revealed that most planets in our interstellar neighborhood are larger than Earth and smaller than Neptune. Collectively termed super-Earths and mini-Neptunes, some of these planets may have the conditions to support liquid water oceans, and thus Earth-like biology, despite differing in many ways from our own planet. In addition to their quantitative abundance, super-Earths are relatively large and are thus more easily detected than true Earth twins. As a result, super-Earths represent a uniquely powerful opportunity to discover and explore a panoply of fascinating and potentially habitable planets in 2020 - 2030 and beyond.

astro-ph.EP↗