127 research outputs found
Overview of the author identification task at PAN 2014
No full textThe author identification task at PAN-2014 focuses on author verification. Similar to PAN-2013 we are given a set of documents by the same author along with exactly one document of questioned authorship, and the task is to determine whether the known and the questioned documents are by the same author or not. In comparison to PAN-2013, a significantly larger corpus was built comprising hundreds of documents in four natural languages (Dutch, English, Greek, and Spanish) and four genres (essays, reviews, novels, opinion articles). In addition, more suitable performance measures are used focusing on the accuracy and the confidence of the predictions as well as the ability of the submitted methods to leave some problems unanswered in case there is great uncertainty. To this end, we adopt the c@1 measure, originally proposed for the question answering task. We received 13 software submissions that were evaluated in the TIRA framework. Analytical evaluation results are presented where one language-independent approach serves as a challenging baseline. Moreover, we continue the successful practice of the PAN labs to examine meta-models based on the combination of all submitted systems. Last but not least, we provide statistical significance tests to demonstrate the important differences between the submitted approaches
Spiritus Loci
No full textIn Spiritus Loci Bert Daelemans, who graduated as an architect and a theologian, provides an interdisciplinary method for the theological assessment of church architecture. Rather than a theory, this method is based on case studies of contemporary buildings (1995-2015), which are often criticized for lacking theological depth. In a threefold method, the author brings to light the ways in which architecture can be theology – or theotopy – by focusing on topoi (places) rather than logoi (words). Churches reveal our relationship with God by engaging our body, mind, and community. This method proves relevant not only for the way we perceive these buildings, but also for the way we use them, especially in our prophetic engagement for a better world
Pregnant women admitted to hospital with covid-19 in 10 European countries: individual patient data meta-analysis of population based cohorts in International Obstetric Survey Systems
No full textObjectives To assess the incidence of hospital admissions for covid-19 disease in pregnant women, severity of covid-19 disease, and medical treatment provided to pregnant women with moderate to severe covid-19 during the first 10 months of the pandemic. Design Individual patient data meta-analysis of population based cohorts in International Obstetric Survey Systems. Setting 10 European countries with national or regional surveillance within the International Obstetric Survey Systems (INOSS) collaboration using aligned definitions and case report forms: Belgium, France (regional), Italy, the Netherlands, Denmark, Finland, Iceland, Norway, Sweden (regional), and the UK. The dominant variant of the SARS-CoV-2 virus was the wild-type variant in all countries during the study period (1 March 2020 to 31 December 2020). Participants The source population was 1.7 million women giving birth (maternities) from 1 March 2020 to 31 December 2020; pregnant women were included if they were admitted to hospital and had a positive polymerase chain reaction test for the SARS-CoV-2 virus ≤7 days before hospital admission, during admission, or up to two days after giving birth. We further categorised the hospital admission in two groups; covid-19 admission (hospital admission due to covid-19 or with reported symptoms of covid-19 disease) or non-covid-19 admission (admission to hospital for obstetric healthcare or no symptoms of covid-19 disease). Main outcome measures Incidence of hospital admissions for covid-19 per 1000 maternities, frequency of moderate to severe covid-19 disease, and number of women who received specific medical treatment for SARS-CoV-2 infection. Moderate to severe covid-19 disease was defined as maternal death, admission to an intensive care unit, or need for respiratory support. Results Among 1.7 million maternities, 9003 women were included in the study: 2350 (26.1%) were admitted to hospital because of covid-19 disease or had symptoms of disease. The pooled incidence of hospital admissions for covid-19 per 1000 maternities was 0.8 (95% confidence interval (CI) 0.5 to 1.2, τ 2 =0.44), ranging from no hospital admissions in Iceland to 1.7 in France and 1.9 in the UK. 13 women died due to covid-19. Among 2219 women admitted to hospital for covid-19 in countries with complete information on respiratory support, 820 women (39.5%, 95% CI 34.6% to 44.4%, τ 2 =0.02) had moderate to severe covid-19 disease. At most, 16.8% (95% CI 7.7% to 32.9%, I 2 =81.8%, τ 2 =0.7) of women with moderate to severe covid-19 received specific medical treatment for SARS-CoV-2 infection with corticosteroids, although 66.6% (59.4% to 73.2%, I 2 =50.1, τ 2 =0.06) were given thromboprophylaxis. Conclusions Population based surveillance in 10 European countries during the first 10 months of the covid-19 pandemic showed variations in the risk of hospital admissions for covid-19 in pregnant women. This finding indicates that national public health policies likely had a substantial and previously unrecognised role in protecting pregnant women. Few pregnant women with moderate to severe covid-19 were given specific medical treatment for SARS-CoV-2 disease, even when there were no or minor safety concerns. Lessons for future pandemics include the importance of rapid, robust surveillance systems for maternal and perinatal health, and of including use for pregnant women early in the development and testing of medicines and vaccines for public health emergencies.info:eu-repo/semantics/publishe
The effect of author set size and data size in authorship attribution
No full textAbstract: Applications of authorship attribution `in the wild [Koppel, M., Schler, J., and Argamon, S. (2010). Authorship attribution in the wild. Language Resources and Evaluation. Advanced Access published January 12, 2010:10.1007/s10579-009-9111-2], for instance in social networks, will likely involve large sets of candidate authors and only limited data per author. In this article, we present the results of a systematic study of two important parameters in supervised machine learning that significantly affect performance in computational authorship attribution: (1) the number of candidate authors (i.e. the number of classes to be learned), and (2) the amount of training data available per candidate author (i.e. the size of the training data). We also investigate the robustness of different types of lexical and linguistic features to the effects of author set size and data size. The approach we take is an operationalization of the standard text categorization model, using memory-based learning for discriminating between the candidate authors. We performed authorship attribution experiments on a set of three benchmark corpora in which the influence of topic could be controlled. The short text fragments of e-mail length present the approach with a true challenge. Results show that, as expected, authorship attribution accuracy deteriorates as the number of candidate authors increases and size of training data decreases, although the machine learning approach continues performing significantly above chance. Some feature types (most notably character n-grams) are robust to changes in author set size and data size, but no robust individual features emerge
Personae: a corpus for author and personality prediction from text
No full textWe present a new corpus for computational stylometry, more specifically authorship attribution and the prediction of author personality from text. Because of the large number of authors (145), the corpus will allow previously impossible studies of variation in features considered predictive for writing style. The innovative meta-information (personality profiles of the authors) associated with these texts allows the study of personality prediction, a not yet very well researched aspect of style. In this paper, we describe the contents of the corpus and show its use in both authorship attribution and personality prediction. We focus on features that have been proven useful in the field of author recognition. Syntactic features like part-of-speech n-grams are generally accepted as not being under the author’s conscious control and therefore providing good clues for predicting gender or authorship. We want to test whether these features are helpful for personality prediction and authorship attribution on a large set of authors. Both tasks are approached as text categorization tasks. First a document representation is constructed based on feature selection from the linguistically analyzed corpus (using the Memory-Based Shallow Parser (MBSP)). These are associated with each of the 145 authors or each of the four components of the Myers-Briggs Type Indicator (Introverted-Extraverted, Sensing-iNtuitive, Thinking-Feeling, Judging-Perceiving). Authorship attribution on 145 authors achieves results around 50 % accuracy. Preliminary results indicate that the first two personality dimensions can be predicted fairly accurately. 1
PAN15 Author Identification: Verification
No full text<p>We provide you with a training corpus that comprises a set of author verification problems in several languages/genres. Each problem consists of some (up to five) known documents by a single person and exactly one questioned document. All documents within a single problem instance will be in the same language. However, their genre and/or topic may differ significantly. The document lengths vary from a few hundred to a few thousand words.</p>
<p>The documents of each problem are located in a separate folder, the name of which (problem ID) encodes the language of the documents. The following list shows the available sub-corpora, including their language, type (cross-genre or cross-topic), code, and examples of problem IDs:</p>
<p>Language; Type; Code; Problem IDs<br>
Dutch; Cross-genre; DU; DU001, DU002, DU003, etc.<br>
English; Cross-topic; EN; EN001, EN002, EN003, etc.<br>
Greek; Cross-topic; GR; GR001, GR002, GR003, etc.<br>
Spanish; Cross-genre; SP; SP001, SP002, SP003, etc.</p>
<p>The ground truth data of the training corpus found in the file <code>truth.txt</code> include one line per problem with problem ID and the correct binary answer (Y means the known and the questioned documents are by the same author and N means the opposite). For example:</p>
<pre>EN001 N
EN002 Y
EN003 N
...</pre>
PAN18 Author Identification: Attribution
No full textWe provide a corpus which comprises a set of cross-domain authorship attribution problems in each of the following 5 languages: English, French, Italian, Polish, and Spanish. Note that we specifically avoid to use the term 'training corpus' because the sets of candidate authors of the development and the evaluation corpora are not overlapping. Therefore, your approach should not be designed to particularly handle the candidate authors of the development corpus.
Each problem consists of a set of known fanfics by each candidate author and a set of unknown fanfics located in separate folders. The file problem-info.json that can be found in the main folder of each problem, shows the name of folder of unknown documents and the list of names of candidate author folders.
The true author of each unknown document can be seen in the file ground-truth.json, also found in the main folder of each problem.
In addition, to handle a collection of such problems, the file collection-info.jsonincludes all relevant information. In more detail, for each problem it lists its main folder, the language (either "en", "fr", "it", "pl", or "sp") and encoding (always UTF-8) of its documents.
More information: Lin
High-throughput analysis of candidate imprinted genes and allele-specific expression
No full textIn diploid human organisms, the ~20,000 genes are usually functional as two active copies or alleles. Exceptionally, some genes have only one active allele while the other is silenced. Two different groups of genes fall into this minor category; the genes that exhibit random monoallelic expression (e.g. odorant receptor genes and genes coding for immunoglobulins), and those genes exhibiting monoallelic expression in a parent-of-origin specific manner, named imprinted genes. At the outset of this study in October 2006, 56 genes in humans were known to be imprinted and 98 in mice, but the total number of imprinted genes in either species was unknown. I have used high-throughput allele-specific PCR assays to screen human term placental tissue samples for new imprinted genes. Hundreds of genes were tested either because they were predicted to be imprinted or because they were candidates for which the imprinted status was simply unknown in human term placenta. My results suggested that we are reaching saturation in the number of human placentally imprinted genes. I show that ZNF331 is imprinted in human placenta and is part of a primate lineagespecific imprinted locus showing differential methylation. My data also highlights that parental allelic specific expression is a continuum, from imprinted monoallelic expression to partial imprinting (i.e., one parental allele is slightly more (or less) expressed than the other). This continuum suggests a requirement to sequence the transcriptome of every human tissue at each different developmental stage exhaustively to assess genes for parent-of-origin specific expression and to clearly define what imprinting means. Most importantly whether ‘partial’ imprinting has functional significance or is just a part of the dynamic flux of gene expression. Specifically, my results call for thorough investigation of the ZNF331 locus in human development, physiology and parent-of-origin specific diseases
PAN14 Author Identification: Verification
No full textWe provide you with a training corpus that comprises a set of author verification problems in several languages/genres. Each problem consists of some (up to five) known documents by a single person and exactly one questioned document. All documents within a single problem instance will be in the same language and best efforts are applied to assure that within-problem documents are matched for genre, register, theme, and date of writing. The document lengths vary from a few hundred to a few thousand words.
More information: Lin
- …
