• seminar

seminar

SEMINAR

In January 2012 Karin Friberg Heppin and Ann-Marie Eklund participated in PROMISE winter school on Information Retrieval meets Information Visualization. In this presentation we will briefly describe the two main topics of the winter school, Information Retrieval and Information Visualization. We will also present a few highlights from the lectures.

http://www.promise-noe.eu/events/winter-school-2012

Date: 2012-02-23 10:15 - 12:00

Location: L308, Lennart torstenssonsgatan 8

Permalink

SEMINAR

Our aim is to detect whether texts written in natural language contain normative conflicts (i.e., whether there are conflicting obligations, permissions and prohibitions). In this talk we present AnaCon, a framework where such texts are written in Controlled Natural Language (CNL) and automatically translated into the formal language CL using the Grammatical Framework (GF). In AnaCon such CL expressions are analyzed for normative conflicts by the tool CLAN which gives a counter-example in case a conflict is found. Our framework uses GF  to give a CNL version of the counter-example helping the user to identify the conflicts in the original text.

The talk will be divided in two parts. In the first part we will introduce AnaCon as well as relevant background on CL and CLAN. In the second part we will we show the usability of AnaCon on two case studies.

Date: 2012-02-16 10:15 - 12:00

Location: EDIT room, EDIT building, Chalmers

Permalink

SEMINAR

Date: 2012-01-26 10:15 - 12:00

Location: L308, Lennart torstenssonsgatan 8

Permalink

SEMINAR

The talk reports recent work with Tom Kwiatkowski, Sharon Goldwater, and Luke Zettlemoyer on semantic parser induction by machine from a number of corpora pairing sentences with logical forms, including GeoQuery and a corpus consisting of real child-directed utterance from the CHILDES corpus.

The problem of semantic parser induction and child language acquisition are both similar to the problem of inducing a grammar and a parsing model from a treebank such as the Penn treebank, except that the trees are unordered logical forms, in which the preterminals are not aligned with words in the target language, and there may be noise and spurious distracting logical forms supported by the context but irrelevant to the utterance.

The talk shows that this class of problem can be solved if the child or machine initially parses with the entire space of possibilities that universal grammar allows under the assumptions of the Combinatory Categorial theory of grammar (CCG), and learns a statistical parsing model for that space using EM-related methods such as Variational Bayes learning.

This can be done without all-or-none "parameter-setting" or attendant "triggers", and without invoking any "subset principle" of the kind proposed in linguistic theory, provided the system is presented with a representative sample of reasonably short string-meaning pairs from the target language.

Mark Steedman on the Web

Date: 2012-02-09 10:15 - 12:00

Location: L308, Lennart torstenssonsgatan 8

Permalink

SEMINAR

Date: 2012-01-09 10:00 - 11:00

Location: The EDIT room at Chalmers

Permalink

SEMINAR

Planning event for the CLT seminar series Spring semester

Date: 2012-01-19 10:15 - 12:00

Location: L308, Lennart torstenssonsgatan 8

Permalink

SEMINAR

I anslutning till projektet Svenskt frasnät (FN++) är vi på väg att utveckla en konstruktionsdatabas, ett konstruktikon, inspirerat av det engelskspråkiga constructicon som är under uppbyggnad vid UC Berkeley som ett tillägg till FrameNet. Tanken är att komplettera befintliga lexikografiska resurser med en uppsättning beskrivningar av strukturer som är för generella för att kunna knytas till enskilda uppslagsord (men för specifika för att betraktas som generella grammatiska regler): konstruktioner. Analyserna bör vara relativt enkla, ungefär motsvarande ordboksdefinitioner, för att möjliggöra beskrivningar i stor skala. Resursen ska helst vara till nytta för såväl språkteknologiska som lingvistiska och pedagogiska ändamål. Vid seminariet den 13/12 diskuteras databasens innehåll och format utifrån några preliminära analysskisser.

Rekommenderad bakgrundsläsning, för den som ev. vill förbereda sig: Fillmore, Charles, Russell Lee-Goldman & Russell Rhodes (to appear). The FrameNet Constructicon. To appear in Boas, Hans and Ivan Sag (eds.), Sign-Based Construction Grammar. Stanford: CSLI. <http://www.icsi.berkeley.edu/pubs/ai/framenetconstructicon11.pdf>

Välkomna!

/Benjamin

Date: 2011-12-13 10:15 - 12:00

Location: L307, Lennart Torstenssonsgatan 8

Permalink

SEMINAR

The web is making all sorts of differences to how we live, work, shop, study, vote and chat. But it seems to change depending on how we look at it.  Sometimes it seems like a library, at others like a shopping mall, at others like a party.  How do companies like Google, Facebook, Microsoft and Apple fit into the picture?  Is every conversation I have on skype 'on the web'?  Is it something that used to be public, but is now turning private?  The question I'll be battling with in the talk is how these different threads fit together.

http://www.kilgarriff.co.uk/

Date: 2011-11-22 13:15 - 15:00

Location: Olof Wijksgatan 6

Permalink

SEMINAR

In the talk I am going to give you an example for an exploratory analysis of the lexical resource SALDO using the tool visone. I will furthermore show some first results of our visual analysis of Web Session Logs.

Date: 2011-10-27 13:15 - 15:00

Location: L307, Lennart torstenssonsgatan 8

Permalink

SEMINAR

How can meaning be compared and evaluated in realistic situations, in which ill-formed language or differences in situative knowledge or world knowledge make a complete analysis difficult or impossible? In this talk I report on our ongoing work in the CoMiC project (SFB833-A4). We are exploring which linguistic representations are effective and robust in a computational-linguistic comparison of the meaning of sentences and text fragments. We focus on learner answers to reading comprehension questions, for which we are collecting large authentic corpora. Using these task-specific corpora, we study the properties of questions and texts which support the comparison of the meaning of answers at different levels of analysis from surface forms to deeper linguistic representations.

 

Professor Detmar Meurers on the web: http://www.sfs.uni-tuebingen.de/~dm/

Date: 2011-12-15 10:15 - 12:00

Location: L308, Lennart Torstenssonsgatan 8

Permalink

X
Loading