Welcome to the seminar on Jan. 20, 2015 at 12:15 in room HF:371.
Vibeke Rønneberg: The need to read – The case of blind writing in dyslexia.
This study investigated the written products and the writing processes of a group of dyslexic students as compared to typically developing students. Research has shown that people with reading and writing difficulties have a strong focus on the writing of single words. We aimed to answer two questions: First, are there differences between dyslexics and typical writers regarding the process and product of writing, and are these associated with a greater word-level focus in dyslexics? Second, are there differences between the groups in how they use visual feedback while writing?
On September 10, Dr. Preslav Nakov from the Qatar Computing Research Institute will give a public master class to Ph.D. candidate Carla Parra at the University of Bergen. The theme will be Nominal compounds in German and their translation to Spanish. This event is organized by the Norwegian Graduate Researcher School in Linguistics and Philology. It will be held from 13:00 to 16:30 in Sydneshaugen skole, room 304B. Interested participants are requested to sign up by email to Bamba Dione by September 8, so that coffee and refreshments can be ordered.
On September 11, Dr. Preslav Nakov will give a lecture entitled The Web as an Implicit Training Set: Application to Noun Compounds Syntax and Semantics. This event will start at 10:15 in HF-bygget, room 216. In this talk, he will reveal some of the hidden potential of the Web that lies beyond the n-gram, with focus on the syntax and semantics of English noun compounds. First he will present a highly accurate lightly supervised approach based on surface markers and linguistically-motivated paraphrases that yields state-of-the-art results for noun compound bracketing: e.g., “[[liver cell] antibody]” is left-bracketed, while “[liver [cell line]]” is right-bracketed. Second, he will present a simple unsupervised method for mining implicit predicates that can characterize the semantic relations holding between the nouns in noun compounds, e.g., “malaria mosquito” is a “mosquito that carries/spreads/causes/transmits/… malaria”. Finally, it will be shown how these ideas can be used to improve statistical machine translation.
On Monday, September 8th, 14:15-15:00 in rom 301, HF-bygget, scientific director Mila Vulchanova will present the Norwegian Graduate Researcher School in Linguistics and Philology (LingPhil). The main aim of this graduate school is to provide young researchers with a broad training in theories and research methods in linguistics and philology by offering a high quality, structured course programme that complements and deepens the training courses offered locally by each of the individual participating institutions. Participants to the meeting will have a chance to ask questions and provide input to the future course portfolio.
Fredagsseminar av Carla Parra og Koenraad De Smedt, med bidrag fra Gunn Inger Lyse (UiB)
Fredag 20. juni kl. 14:15 i rom HF:216
Av mange grunner er det ønskelig at forskingsresultater i form av språkressurser arkiveres og gjøres gjenbrukbare. God forvaltling av språkressurser gjør data søkbar og siterbar, gjør forskningen gjentagbar, og gjør videre forskning sammenlignbar med annen forskning på samme materiale. Gjenbruk har en rekke forutsetninger, bl.a. at det lages god dokumentasjon og kompatible metadata, men også at juridiske og etiske forhold blir avklart. Lisensiering av språkressurser har også den spesielle utfordringen at tekstkildene ofte er rettighetsbelagt av tredjepart. Dette seminaret gir en kort innføring i problemstillingen rundt copyright og mulighetene for å lisensiere språkressurser, bl.a. gjennom standardlisenser.
Friday May 23, 2014 at 14:15 – 16:00 (HF: 217)
Robert Paulsen (PhD researcher, Research Group for Medieval Philology)
Old Norwegian is unique among the Germanic languages in featuring a vowel harmony. Vowel harmony is a phonological process that may be described as a kind of reverse umlaut: The quality of a vowel in a following syllable is influenced by the quality of a preceding vowel. For Old Norwegian this means that the unstressed vowels /i/ and /u/ (/a/ does not participate in vowel harmony) in derivational and inflectional suffixes are partially assimilated (with regard to tongue height) to the preceding stem vowel. Thus, in manuscripts unstressed /i/ and /u/ will be represented by either <i>, <u> or by <e>, <o> – depending on the tongue height of the preceding vowel.
The situation, however, is somewhat more complicated than this: with some vowels, the vowel harmony process appears to contradict what we know about the relative tongue height of Old Norwegian vowels, raising questions about both the workings of vowel harmony in general and about our understanding of Old Norse phonology.
In my presentation, I will discuss the Old Norwegian vowel harmony – both as it is traditionally described and as I have observed it in one particular manuscript. This leading to typological questions I will compare Old Norwegian vowel harmony to similar features in languages spoken today and try to make sense out of the apparently problematic phonology behind this strikingly unique phenomenon.
Fredag 11. april 2014, 14:15 – 16:00 (HF: 217)
Victoria Rosén, Helge Dyvik, Petter Haugereid, Martha Thunes og Gyri Smørdal Losnegaard, forskergruppe LaMoRe (Language Models and Resources), UiB
INESS (Infrastructure for the Exploration of Syntax and Semantics) er et prosjekt i NFRs INFRASTRUKTUR-program. Prosjektet bygger en infrastruktur for trebanker, som er syntaktisk annoterte korpora. Annotasjonen gjør det mulig å finne grammatiske konstruksjoner som det er vanskelig å finne i korpora uten syntaktisk annotasjon. Et avansert søkesystem gjør det mulig for grammatikkforskere, leksikografer og andre interesserte å søke etter eksempler i trebanken på en detaljert og fleksibel måte.
INESS-prosjektet har to hovedmål: å gjøre trebanker for mange språk søkbare og synlige gjennom en vanlig nettleser, og å bygge en omfattende norsk trebank med detaljerte syntaktiske og funksjonelle analyser for bokmål og nynorsk. I dette seminarinnlegget vil noen av INESS-medarbeiderne presentere viktige deler av prosjektet, og da særlig hvordan den norske trebanken bygges. Blant annet vil de vise hvordan de syntaktiske analysene ser ut, hvordan tekster preprosesseres før de analyseres automatisk, og hvordan grammatikk og leksikon revideres i møte med korpusdata.
Ifølge videnskab.dk skal et nytt dansk forskningsprosjekt lære å forstå dansk særlig ved å skille mellom betydninger av flertydige ord.
|
|