2,789 research outputs found
Data and models for metonymy resolution
No full textWe describe the first shared task for figurative language resolution, which was organised within SemEval-2007 and focused on metonymy. The paper motivates the linguistic principles of data sampling and annotation and shows the task’s feasibility via human agreement. The five participating systems mainly used supervised approaches exploiting a variety of features, of which grammatical relations proved to be the most useful. We compare the systems’ performance to automatic baselines as well as to a manually simulated approach based on selectional restriction violations, showing some limitations of this more traditional approach to metonymy recognition. The main problem supervised systems encountered is data sparseness, since metonymies in general tend to occur more rarely than literal uses. Also, within metonymies, the reading distribution is skewed towards a few frequent metonymy types. Future task developments should focus on addressing this issue
SemEval-2007 Task 08: Metonymy Resolution at SemEval-2007
No full textTask description for the Metonymy Resolution Task
Learning to buy a Renault and talk to BMW: A supervised approach to conventional metonymy
No full textWe test the validity of a machine learning approach to metonymy
resolution, presenting a case study for organisation names and comparing the results with a previous experiment on location names. We de-scribe a reliable annotation scheme for organisation names and present a corpus annotated for metonymic usage. We then discuss experiments with a supervised classification algorithm on this corpus, focusing on feature contributions and highlighting advantages and disadvantages of a classification approach to metonymy resolution
Metonymic Proper Names: A Corpus-based Account
No full textMany proper names are widely used metonymically. Thus, for example, organisation names
can be used for products produced by the organisation, members of an organisation or events
associated with the organisation. Their treatment is crucial for many natural language processing
tasks like question answering and anaphora resolution. At the moment, language resources do not
contain the necessary information for large-scale metonymy processing.
As a contribution, we describe a general framework for annotating metonymies in domain-
independent text that considers the regularity, productivity and underspecification of metonymic
usage. We will then concentrate on two fully worked out annotation schemes for location and
organisation names and rigorously evaluate these schemes as to their reliability. We also present a
gold standard corpus consisting of 4000 annotated occurrences of location and organisation names
in the British National Corpus. We use this corpus to examine the distribution of metonymies as
well as for experiments in automatic metonymy resolution
Comparing knowledge sources for nominal anaphora resolution
No full textWe compare two ways of obtaining lexical knowledge for antecedent selection in other-anaphora
and definite noun phrase coreference. Specifically, we compare an algorithm that relies on links
encoded in the manually created lexical hierarchy WordNet and an algorithm that mines corpora
by means of shallow lexico-semantic patterns. As corpora we use the British National
Corpus (BNC), as well as the Web, which has not been previously used for this task. Our
results show that (a) the knowledge encoded in WordNet is often insufficient, especially for
anaphor-antecedent relations that exploit subjective or context-dependent knowledge; (b) for
other-anaphora, the Web-based method outperforms the WordNet-based method; (c) for definite
NP coreference, the Web-based method yields results comparable to those obtained using
WordNet over the whole dataset and outperforms the WordNet-based method on subsets of the
dataset; (d) in both case studies, the BNC-based method is worse than the other methods because
of data sparseness. Thus, in our studies, the Web-based method alleviated the lexical knowledge
gap often encountered in anaphora resolution, and handled examples with context-dependent relations
between anaphor and antecedent. Because it is inexpensive and needs no hand-modelling
of lexical knowledge, it is a promising knowledge source to integrate in anaphora resolution systems
Data for: On the merging of optical and SAR satellite imagery for surface water mapping applications
No full textData on the optical-SAR data fusion for surface water mapping. "sar-optical-timeseries.xlsx" is the time series data with two sheets, one being the SAR timeseries and the other being the optical time series. "hydraCase_Nov_sar.tif" is a 2-band Geotiff file of the water extraction methods on SAR data with the first band being the water index and the second band being the extracted water. "hydraCase_nov_opt.tif" is the same as "hydraCase_Nov_sar.tif" but for the optical sensor
Numerical analysis of accidental hydrogen releases from high pressure storage at low temperatures
No full textEvaluations of the performance of simplified engineering and CFD models are important to improve risk assessment tools e.g. to predict accurately releases from various types of hydrogen storages. These tools have to predict releases from a wide range of storage pressures (up to 80 MPa) and temperatures (down to 20 K), e.g. cryogenic compressed gas storage covers pressures up to 35 MPa and temperatures between 33 K and 338 K. Accurate calculations of high pressure releases require real gas EOS. This paper compares a number of EOS to predict hydrogen properties typical in different storage types. The vessel dynamics are modeled using a simplified engineering and a CFD model to evaluate the performance of various EOS to predict vessel pressures, temperatures mass flow rates and jet flame lengths. It is shown that the chosen EOS and the chosen specific heat capacity correlation are important to model accurately hydrogen releases at low temperatures
Overvåking av miljøgifter i sedimentene i Grenlandsfjordene 1997
No full textSom del av overvåkingen av miljøtilstanden i Grenlandsfjordene ble sedimentene prøvetatt i 1997. Resultatene viser at de klorerte forbindelsene som har Hydro Porsgrunn som hovedkilde (dioksiner, heksaklorbenzen, etc) generelt var redusert til under det halve av konsentrasjonene i 1989 i samsvar med de store reduksjonene i utslippene. For kvikksølv har det vært små forandringer etter at direktetilførsler stoppet i 1987. Det var også små endringer i konsentrasjonen av PAH. Større endringer var forventet utfra reduksjonene i utslippene fra Elkem Mangan KS PEA som har vært den tidligere antatte hovedkilden. Selv om konsentrasjonene har avtatt, er sedimentene fremdeles tildels betydelig forurenset. I henhold til SFTs kvalitetskriterier, klassifiseres sedimentene som markert til meget sterkt forurenset med hensyn på dioksiner, sum DDT og PAH. Sedimentene er generelt lite påvirket til markert påvirket av PCB og kvikksølv. Det ble også observert en betydelig påvirkning med polyklorerte naftalener med en 20-gangers forskjell i konsentrasjon mellom Frierfjorden og ytre deler av Brevikfjorden
- …
