Named Entity Recognition (NER) is a key task to support biomedical research. In Biomedical Named Entity Recognition (BioNER), obtaining high-quality expert annotated data is laborious and expensive, leading to the development of automatic approaches such as distant supervision. However, manually and automatically generated data often suffer from the unlabeled entity problem, whereby many entity annotations are missing, degrading the performance of full annotation NER models. To conquer this issue, we undertake a systematic exploration of the efficacy of partial annotation learning methods for BioNER, which encompasses a comprehensive evaluation conducted across a spectrum of distinct simulated scenarios of missing entity annotations. Furthermore, we propose a TS-PubMedBERT-Partial-CRF partial annotation learning model. We standardize a compilation of 16 BioNER corpora, encompassing a range of five distinct entity types, to establish a gold standard. And we compare against the state-of-the-art partial annotation model EER-PubMedBERT, the widely acknowledged partial annotation model BiLSTM-Partial-CRF model, and the state-of-the-art full annotation learning BioNER model PubMedBERT tagger. Results show that partial annotation learning-based methods can effectively learn from biomedical corpora with missing entity annotations. Our proposed model outperforms alternatives and, specifically, the PubMedBERT tagger by 38% in F1-score under high missing entity rates. Moreover, the recall of entity mentions in our model demonstrates a competitive alignment with the upper threshold observed on the fully annotated dataset.

Abstract We propose a decentralised “local2global” approach to graph representation learning, that one can a-priori use to scale any embedding technique. Our local2global approach proceeds by first dividing the input graph into overlapping subgraphs (or “patches”) and training local representations for each patch independently. In a second step, we combine the local representations into a globally consistent representation by estimating the set of rigid motions that best align the local representations using information from the patch overlaps, via group synchronization. A key distinguishing feature of local2global relative to existing work is that patches are trained independently without the need for the often costly parameter synchronization during distributed training. This allows local2global to scale to large-scale industrial applications, where the input graph may not even fit into memory and may be stored in a distributed manner. We apply local2global on data sets of different sizes and show that our approach achieves a good trade-off between scale and accuracy on edge reconstruction and semi-supervised classification. We also consider the downstream task of anomaly detection and show how one can use local2global to highlight anomalies in cybersecurity networks.

Neural language models are the backbone of modern-day natural language processing applications. Their use on textual heritage collections which have undergone Optical Character Recognition (OCR) is therefore also increasing. Nevertheless, our understanding of the impact OCR noise could have on language models is still limited. We perform an assessment of the impact OCR noise has on a variety of language models, using data in Dutch, English, French and German. We find that OCR noise poses a significant obstacle to language modelling, with language models increasingly diverging from their noiseless targets as OCR quality lowers. In the presence of small corpora, simpler models including PPMI and Word2Vec consistently outperform transformer-based models in this respect.

By linking to external websites, Wikipedia can act as a gateway to the Web. To date, however, little is known about the amount of traffic generated by Wikipedia’s external links. We fill this gap in a detailed analysis of usage logs gathered from Wikipedia users’ client devices. Our analysis proceeds in three steps: First, we quantify the level of engagement with external links, finding that, in one month, English Wikipedia generated 43M clicks to external websites, in roughly even parts via links in infoboxes, cited references, and article bodies. Official links listed in infoboxes have by far the highest click-through rate (CTR), 2.47% on average. In particular, official links associated with articles about businesses, educational institutions, and websites have the highest CTR, whereas official links associated with articles about geographical content, television, and music have the lowest CTR. Second, we investigate patterns of engagement with external links, finding that Wikipedia frequently serves as a stepping stone between search engines and third-party websites, effectively fulfilling information needs that search engines do not meet. Third, we quantify the hypothetical economic value of the clicks received by external websites from English Wikipedia, by estimating that the respective website owners would need to pay a total of $7–13 million per month to obtain the same volume of traffic via sponsored search. Overall, these findings shed light on Wikipedia’s role not only as an important source of information, but also as a high-traffic gateway to the broader Web ecosystem.

A growing volume of heritage data is being digitized and made available as text via optical character recognition (OCR). Scholars and libraries are increasingly using OCR-generated text for retrieval and analysis. However, the process of creating text through OCR introduces varying degrees of error to the text. The impact of these errors on natural language processing (NLP) tasks has only been partially studied. We perform a series of extrinsic assessment tasks — sentence segmentation, named entity recognition, dependency parsing, information retrieval, topic modelling and neural language model fine-tuning — using popular, out-of-the-box tools in order to quantify the impact of OCR quality on these tasks. We find a consistent impact resulting from OCR errors on our downstream tasks with some tasks more irredeemably harmed by OCR errors. Based on these results, we offer some preliminary guidelines for working with text produced through OCR.

A variety of schemas and ontologies are currently used for the machine-readable description of bibliographic entities and citations. This diversity, and the reuse of the same ontology terms with different nuances, generates inconsistencies in data. Adoption of a single data model would facilitate data integration tasks regardless of the data supplier or context application. In this paper we present the OpenCitations Data Model (OCDM), a generic data model for describing bibliographic entities and citations, developed using Semantic Web technologies. We also evaluate the effective reusability of OCDM according to ontology evaluation practices, mention existing users of OCDM, and discuss the use and impact of OCDM in the wider open science community.

Wikipedia is one of the most visited sites on the Web and a common source of information for many users. As an encyclopedia, Wikipedia was not conceived as a source of original information, but 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 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’ interactions with citations. 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 deepen our understanding of Wikipedia’s role in a global information economy where reliability is ever less certain, and source attribution ever more vital.

This work presents defoe, a new scalable and portable digital eScience toolbox that enables historical research. It allows for running text mining queries across large datasets, such as historical newspapers and books in parallel via Apache Spark. It handles queries against collections that comprise several XML schemas and physical representations. The proposed tool has been successfully evaluated using five different large-scale historical text datasets and two HPC environments, as well as on desktops. Results shows that defoe allows researchers to query multiple datasets in parallel from a single command-line interface and in a consistent way, without any HPC environment-specific requirements.

We investigated the similarities of pairs of articles that are cocited at the different cocitation levels of the journal, article, section, paragraph, sentence, and bracket. Our results indicate that textual similarity, intellectual overlap (shared references), author overlap (shared authors), proximity in publication time all rise monotonically as the cocitation level gets lower (from journal to bracket). While the main gain in similarity happens when moving from journal to article cocitation, all level changes entail an increase in similarity, especially section to paragraph and paragraph to sentence/bracket levels. We compared the results from four journals over the years 2010–2015: Cell, the European Journal of Operational Research, Physics Letters B, and Research Policy, with consistent general outcomes and some interesting differences. Our findings motivate the use of granular cocitation information as defined by meaningful units of text, with implications for, among others, the elaboration of maps of science and the retrieval of scholarly literature.