To save content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about saving content to .
To save content items to your Kindle, first ensure email@example.com
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about saving to your Kindle.
Note you can select to save to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
We compare two frameworks for the segmentation of words in child-directed speech, PHOCUS and MULTICUE. PHOCUS is driven by lexical recognition, whereas MULTICUE combines sub-lexical properties to make boundary decisions, representing differing views of speech processing. We replicate these frameworks, perform novel benchmarking and confirm that both achieve competitive results. We develop a new framework for segmentation, the DYnamic Programming MULTIple-cue framework (DYMULTI), which combines the strengths of PHOCUS and MULTICUE by considering both sub-lexical and lexical cues when making boundary decisions. DYMULTI achieves state-of-the-art results and outperforms PHOCUS and MULTICUE on 15 of 26 languages in a cross-lingual experiment. As a model built on psycholinguistic principles, this validates DYMULTI as a robust model for speech segmentation and a contribution to the understanding of language acquisition.
Hate speech is any kind of communication that attacks a person or a group based on their characteristics, such as gender, religion and race. Due to the availability of online platforms where people can express their (hateful) opinions, the amount of hate speech is steadily increasing that often leads to offline hate crimes. This paper focuses on understanding and detecting hate speech in underground hacking and extremist forums where cybercriminals and extremists, respectively, communicate with each other, and some of them are associated with criminal activity. Moreover, due to the lengthy posts, it would be beneficial to identify the specific span of text containing hateful content in order to assist site moderators with the removal of hate speech. This paper describes a hate speech dataset composed of posts extracted from HackForums, an online hacking forum, and Stormfront and Incels.co, two extremist forums. We combined our dataset with a Twitter hate speech dataset to train a multi-platform classifier. Our evaluation shows that a classifier trained on multiple sources of data does not always improve the performance compared to a mono-platform classifier. Finally, this is the first work on extracting hate speech spans from longer texts. The paper fine-tunes BERT (Bidirectional Encoder Representations from Transformers) and adopts two approaches – span prediction and sequence labelling. Both approaches successfully extract hateful spans and achieve an F1-score of at least 69%.
Fibrinolysis is an acceptable treatment for acute ST-segment elevation myocardial infarction (STEMI) when primary percutaneous coronary intervention (PCI) cannot be performed within 120 minutes. The American Heart Association has recommended Emergency Medical Services (EMS) interventions such as prehospital fibrinolysis (PHF), prehospital electrocardiogram (ECG), and hospital bypass direct to PCI center. Nova Scotia, Canada has incorporated these interventions into a unique province-wide approach to STEMI care. A retrospective cohort analysis comparing the primary outcome of 30-day mortality for patients receiving either prehospital or emergency department (ED) fibrinolysis (EDF) to patients transported directly by EMS from community or regional ED for primary PCI was conducted.
This retrospective, population-based cohort study included all STEMI patients in Nova Scotia who survived to hospital admission from July 2011 through July 2013. Three provincial databases were used to collect demographic, 30-day mortality, hospital readmission, and rescue PCI data. The results were grouped and compared according to reperfusion strategy received: PHF, EDF, patients brought by ambulance via EMS direct to PCI (EMS to PCI), and ED to PCI (ED to PCI).
There were 1,071 STEMI patients included with 145 PHF, 606 EDF, 98 EMS to PCI, and 222 ED to PCI. There were no significant differences in 30-day mortality across groups (n, %): PHF 5(3); EDF 36(6); EHS to PCI <5(2); and ED to PCI 10(4); P = .28. There was no significant difference in patients receiving fibrinolysis who underwent rescue PCI.
Prehospital fibrinolysis incorporated into a province-wide approach to STEMI treatment is feasible with no observed difference in patient 30-day mortality outcomes observed.
We select three word segmentation models with psycholinguistic foundations – transitional probabilities, the diphone-based segmenter, and PUDDLE – which track phoneme co-occurrence and positional frequencies in input strings, and in the case of PUDDLE build lexical and diphone inventories. The models are evaluated on caregiver utterances in 132 CHILDES corpora representing 28 languages and 11.9 m words. PUDDLE shows the best performance overall, albeit with wide cross-linguistic variation. We explore the reasons for this variation, fitting regression models to performance scores with linguistic properties which capture lexico-phonological characteristics of the input: word length, utterance length, diversity in the lexicon, the frequency of one-word utterances, the regularity of phoneme patterns at word boundaries, and the distribution of diphones in each language. These properties together explain four-tenths of the observed variation in segmentation performance, a strong outcome and a solid foundation for studying further variables which make the segmentation task difficult.
Elastic fibres are insoluble components of the extracellular matrix of dynamic connective tissues such as skin, arteries, lungs and ligaments. They are laid down during development, and comprise a cross-linked elastin core within a template of fibrillin-based microfibrils. Their function is to endow tissues with the property of elastic recoil, and they also regulate the bioavailability of transforming growth factor β. Severe heritable elastic fibre diseases are caused by mutations in elastic fibre components; for example, mutations in elastin cause supravalvular aortic stenosis and autosomal dominant cutis laxa, mutations in fibrillin-1 cause Marfan syndrome and Weill–Marchesani syndrome, and mutations in fibulins-4 and -5 cause autosomal recessive cutis laxa. Acquired elastic fibre defects include dermal elastosis, whereas inflammatory damage to fibres contributes to pathologies such as pulmonary emphysema and vascular disease. This review outlines the latest understanding of the composition and assembly of elastic fibres, and describes elastic fibre diseases and current therapeutic approaches.
Long-term care (LTC) patients are often sent to emergency departments (EDs) by ambulance. In this novel extended care paramedic (ECP) program, specially trained paramedics manage LTC patients on site. The objective of this pilot study was to describe the dispatch and disposition of LTC patients treated by ECPs and emergency paramedics.
Data were collected from consecutive calls to 15 participating LTC facilities for 3 months. Dispatch determinants, transport rates, and relapse rates were described for LTC patients attended by ECPs or emergency paramedics. ECP involvement in end-of-life care was identified.
Of 238 eligible calls, 140 (59%) were attended by an ECP and 98 (41%) by emergency paramedics. Although the top three determinants were the same in each group, the overall distribution of dispatch determinants and acuity differed. In the ECP cohort, 98 of 140 (70%) were treated and released, 33 of 140 (24%) had “facilitated transfer” arranged by an ECP, and 9 of 140 (6%) were immediately transported to the ED by ambulance. In the emergency paramedic cohort, 77 of 98 (79%) were immediately transported to the ED and 21 of 98 (21%) were not transported. In the ECP group, 6 of 98 (6%) patients not transported triggered a 911 call within 48 hours for a related clinical reason, although none of the patients not transported by emergency paramedics relapsed.
ECP involvement in LTC calls was found to reduce transports to the ED with a low rate of relapse. These pilot data generated hypotheses for future study, including determination of appropriate populations for ECP care and analysis of appropriate and safe nontransport.
During the Late Roman Empire Terence was the most revered and the most quoted classical Latin poet after Virgil. Among authors both pagan and Christian, none (to judge, of course, by the surviving written record) made as frequent or as creative literary use of his comedies as Jerome, one of the most accomplished polymaths in all of Latin antiquity. In his estimation Terence ranked, alongside Homer, Menander and Virgil, as one of the greatest of all poets. Jerome had an encyclopedic knowledge of Terence's dramatic corpus and quoted or appropriated phraseology from all six of his comedies. A significant number of these reminiscences have already been identified, but others await discovery. The purpose of the present study is to make a further contribution to this particular branch of Hieronymian Quellenforschung by adducing and analysing two hitherto unrecognized allusions in Jerome's correspondence to Terence's Eunuchus, apparently one of the biggest blockbusters in the history of the Roman stage.
Two studies investigating young readers' use of conjunctions are reported. In Study One, 145 eight- to ten-year-olds completed one of two narrative cloze tasks in which different types of conjunction were deleted. Performance for additive conjunctions was not affected by age in this study, but older children were more likely to select the target conjunction than were younger children for temporal, causal, and adversative terms. Performance was superior in the cloze task in which they were given a restricted choice of responses (three vs. seven). In Study Two, 35 eight- and nine-year-old good and poor comprehenders completed the three-choice cloze task. The poor comprehenders were less likely to select the target terms in general. Sentence-level comprehension skills did not account for their poor performance. The results indicate that understanding of the semantic relations expressed by conjunctions is still developing long after these terms are used correctly in children's speech. The findings are discussed in relation to the role of conjunctions in text comprehension.