Results 71 to 80 of about 4,321 (229)

Social media, high‐frequency trading, and market making after‐hours – Evidence from presidential tweets

Journal of Financial Research, EarlyView.
Abstract I analyze differences between the core and extended trading sessions in the high‐frequency reaction of equity markets to potential news. Using presidential tweets as unanticipated, potentially market‐stirring events, I find that volatility increases and liquidity deteriorates within fractions of a second after a tweet.
Stefan Scharnowski
wiley   +1 more source

Web-Application for the Presentation of Bilingual Corpora (Focusing on Bulgarian as One of the Two Paired Languages)

Cognitive Studies | Études cognitives, 2015
Web-Application for the Presentation of Bilingual Corpora (Focusing on Bulgarian as One of the Two Paired Languages) This paper briefly presents a web-application for the presentation of bilingual aligned corpora focusing on Bulgarian as one the two ...
Ludmila Dimitrova, Ralitsa Dutsova
doaj   +1 more source

Lemmatization of Reflexives in Northern Sotho

Lexikos, 2013
The aim of this article is to evaluate current strategies in the lemmatization of reflexives in Northern Sotho. In particular the so-called "traditional" approach according to which reflexives are lemmatized randomly, as well as the more "rule orientated" alternative, will be critically evaluated mainly against the background of principles such as user
openaire   +3 more sources

To lemmatize or not to lemmatize: how word normalisation affects ELMo performance in word sense disambiguation

CoRR, 2019
We critically evaluate the widespread assumption that deep learning NLP models do not require lemmatized input. To test this, we trained versions of contextualised word embedding ELMo models on raw tokenized corpora and on the corpora with word tokens replaced by their lemmas.
Kutuzov, Andrey, Kuzmenko, Elizaveta
openaire   +3 more sources

Code and Creed: The Construction of AI‐Islamic Discourse in Singapore's Media Landscape

Journal for the Scientific Study of Religion, EarlyView.
ABSTRACT This study examines how Singapore's mainstream media shapes public understanding of artificial intelligence (AI) in Islamic contexts through analysis of newspaper coverage from 1989 to 2024. Drawing on computational analysis of over 620,000 articles from The Straits Times and The Business Times, we develop the concept of “double mediation” to ...
Reza Shaker
wiley   +1 more source

Stemming and lemmatization: A comparison of retrieval performances [PDF]

, 2014
The current study proposes to compare document retrieval precision performances based on language modeling techniques, particularly stemming and lemmatization.
Lloyd-Yemoh, Ethel, Balakrishnan, Vimala
core  

Bridging the gap between digital humanities and natural language processing: Text cleaning and NLP accuracy evaluation of a sample of 20th century Romanian novels

Swedish Journal of Romanian Studies
Although Digital Humanities (DH) and Natural Language Processing (NLP) share common ground, the two academic communities around these disciplines rarely benefit from each other.
Bogdan Vătavu, David Morariu
doaj   +1 more source

Disagreement About Fiscal Policy

Oxford Bulletin of Economics and Statistics, EarlyView.
ABSTRACT Politicians disagree about fiscal policy. This disagreement should have economic effects beyond the effects of government spending and taxation. We use the full set of speeches in the German Bundestag since 1960 and apply state‐of‐the art natural language processing techniques to construct two series of fiscal disagreement starting in 1970 ...
Albina Latifi, Viktoriia Naboka‐Krell, Peter Tillmann, Peter Winker   +3 more
wiley   +1 more source

Lexicon-assisted tagging and lemmatization in Latin: A comparison of six taggers and two lemmatization models [PDF]

Proceedings of the 9th SIGHUM Workshop on Language Technology for Cultural Heritage, Social Sciences, and Humanities (LaTeCH), 2015
We present a survey of tagging accuracies — concerning part-of-speech and full morphological tagging — for several taggers based on a corpus for medieval church Latin (see www.comphistsem.org). The best tagger in our sample, Lapos, has a PoS tagging accuracy of close to 96% and an overall tagging accuracy (including full morphological tagging) of about
Tim vor der Brück, Steffen Eger, Alexander Mehler   +2 more
openaire   +1 more source

Community‐Oriented or Self‐Interested? Citizen Motivations for Engaging in Digital Coproduction

Public Administration, EarlyView.
ABSTRACT Public organizations increasingly use digitally enabled platforms such as 311 (a nonemergency service in the United States) to facilitate coproduction and improve service delivery. While prior research has examined coproducers' motivations, we know little about motivations for digital coproduction.
Vishal Trehan, Gregory Porumbescu, Suzanne Piotrowski   +2 more
wiley   +1 more source