LiLa at UiB

How productive is Machine Translation? (Friday seminar)

Carla Parra Escartín (Hermes, Spain): Machine Translation Evaluation and Productivity Thresholds for Machine Translation Post-Editing Tasks

Time and place: April 22, 2016 at 14:15 in HF:435

Abstract: Machine Translation has become a reality in the translation industry. Over the past few years, translators have experienced the introduction of Machine Translation Post-Editing tasks in their workflows, However, the question of whether MT output has a positive impact in productivity is still open. In an experiment, the productivity of ten professional translators was measured when translating from scratch, post-editing Translation Memory fuzzy matches and post-editing Machine Translation output. The results and the productivity thresholds will be discussed.

Short bio: Carla Parra Escartín obtained her PhD in Computational Linguistics at the University of Bergen and is currently a post-doctoral researcher at Hermes Traducciones y Servicios Lingüísticos, a Spanish translation company. She works as Experienced Researcher within the EXPERT ITN. To see her work within the EXPERT project, visit: http://expert-itn.eu/?q=ER2

Linguistic anomaly and regularity

Lecture by Dr. Carl Vogel, Trinity College Dublin

Monday, Jan. 11, 2016 at 13:15, HF:400

Abstract:

Linguistic systems are interesting to study in part because of their regularity: people often find beauty in regularity. However, linguistic anomaly is also worthy of attention. Firstly, as diversions from the norm, anomalies are inherently curiosities. Secondly, some anomalies are adopted as new norms (whether originating as apt novel metaphor, insightful generalization, or some other type of creative formulation — even if not intended as such by the speaker, but recognized as such by others). Thirdly, anomalies that are not taken up by others bring regularities into relief, supporting deeper insight into the regularities.

Focus on distinctive linguistic events has a wide range of applications. For example, sometimes distinctive linguistic events are reports of changes in experience that should be attended to: this is the case when monitoring self-rated health reports of individuals with chronic illness. However, sometimes such distinctive linguistic events are not reports, but symptoms of cognitive decline. This contrast points to a fundamental problem requiring complementary approaches to achieve solution: the discrimination of change in language of an individual (ideolect) from change in language in general (dialect), on one hand, and from change in the underlying described reality, on the other hand. Pursuit of solutions to this problem leads to a range of empirical and analytical explorations: models of language evolution, models of dialogue interaction, narrative analysis informing predictive models of personal health, identification of emergent leaders in social media fora, compuational stylistics applied within the digital humanities, and more.

Progress so far and future directions within this research programme are outlined.

Teenage girls and language change

New research suggests that teenage girls are an influential group that drives language change. If that is true, sociolinguists may need to pay more attention to language processes in this group.

Is Norwegian weird?

Idibon has an article that attempts to identify linguistic outliers based on their characteristics as documented in the World Atlas of Language Structures. According to this article, Norwegian is not really a middle-of-the-road language.

Radikale vs moderate ord

Anja Zawadska Persvold presenterer sin masteroppgave i datalingvistikk og språkteknologi:

Subnormvariasjon i bokmål: en analysator

19. juni 2015 kl. 14:00 i HF:301

En master med stil

Victoria Troland presenterer sin masteroppgave:

Hvem er forfatteren? – Stilometriske undersøkelser av norske prosatekster

12. juni 2015 kl. 11:00 i HF:301

The newswriter objectivity utopia

The POLAME projects presents a lecture on Wed. May 27, 2015 at 14:15 in HF:400

The Newswriter Objectivity Utopia: Exploring Features for Sentiment Analysis on News Articles

by Diego A. Burgos, Department of Romance Languages, Wake Forest University, USA

Is it an editorial I am reading? – Wait! It’s the newspaper’s business section! Have you ever ended up feeling supportive or opposing of public figures or policies after reading the news? Such scenario of persuasiveness is not uncommon in natural languages due to the rhetoric effect of some texts. However, persuasive is not how one would normally tag news articles that are claimed to be objective such as those in the economy or politics sections of dailies. This work explores the discursive and lexico-grammatical features that unveil the newswriter’s subjectivity, if any. We also have the machine learn these features and predict the newswriter’s position, just as a human would do. The results are interesting and shed light on opinion trends in news articles suggesting what could be seen as the newswriter objectivity utopia.