Högskolan i Skövde

How can corpora assist translators in ways in which resources like translation memories or term databases cannot? Our tests on English, Polish and Swedish parts of the JRC-Acquis Multilingual Parallel show that corpora can provide support for term standardization and variation, and, most importantly, for tracing novel expressions. A corpus tool with an explicit dictionary representation is particularly suitable for the last task. Culler is a tool which allows one to select expressions with words absent from its dictionary. Even if the extracted material may be stained with some noise, it has an undeniable value for translators and lexicographers. The quality of extraction depends in a rather obvious way on the dictionary and text processing but also on the query.

To what extent do Modern Greek, Polish, Swedish and American English similarly lexicalize action concepts, and how similar are the semantic associations between verbs denoting natural actions? Previous results indicate cross-linguistic stability between American English, Swedish, and Polish in verbs denoting basic human body movement, mouth movements, and sound production. The research reported here extends the cross-linguistic comparison to include Greek, which, unlike Polish, American English and Swedish, is a path-language. We used action imagery criteria to obtain lists of verbs from native Greek speakers. The data were analyzed by using multidimensional scaling, and the results were compared to those previously obtained.

The dynamic development of information fusion research implies introduction of new terms and concepts, which in turn requires tools and methods for terminology organization and standardization, as well as tools for creating domain-specific ontology. In this paper, we show how natural language processing and corpus technology tools applied for term extraction from texts in biomedicine can successfully be used for the field of information fusion. We demonstrate term and information extraction from a corpus of research articles in information fusion, showing how a vision of a combined text retrieval and information extraction service can be made real.

In this study, the use of most frequent spatial prepositions in English, Polish, Swedish, and Russian is analyzed. The prepositions and their contexts are extracted from corpora by means of concordance tools. The collostructional strength between the prepositions and the most frequent nouns in the PPs (Gries et al. 2005) is then computed in order to get a more detailed picture of the contexts in which a given preposition is most likely to appear. The results of the investigation are then analysed within the framework of cognitive semantics, especially Croft and Cruse's (2004) taxonomy of construal operations, and Talmy’s (2005) classification of spatial images

Motivation: Automated text analysis is an important tool for facilitating the extraction of knowledge from biomedical abstracts, thereby enabling researchers to build pathway models that integrate and summarize information from a large number of sources. Advanced methods of in-depth analysis of texts using grammar-based approaches developed within the field of computational linguistics must be adapted to the special requirements and challenges posed by biomedical texts, so that these methods can be made available to the bioinformatics and computational biology communities. Results: Our system for automated text analysis and extraction of pathway information is here applied to a set of PubMed abstracts concerning the CBF signaling pathway, which is a key pathway involved in the cold-adaptation response of plants subjected to cold non-freezing temperatures. The system successfully and accurately re-discovers the main features of this pathway, while also pointing to interesting and plausible new hypotheses. The evaluation also reveals a number of issues which will be important targets in the continued development of the system, e.g. the need for an extended lexicon of taxonomic terms and an improved procedure for recognition of sentence boundaries.

We demonstrate how automated text analysis can be used to support the large-scale analysis of metabolic and regulatory pathways by deriving pathway maps from textual descriptions found in the scientific literature. The main assumption is that correct syntactic analysis combined with domain-specific heuristics provides a good basis for relation extraction. Our method uses an algorithm that searches through the syntactic trees produced by a parser based on a Referent Grammar formalism, identifies relations mentioned in the sentence, and classifies them with respect to their semantic class and epistemic status (facts, counterfactuals, hypotheses). The semantic categories used in the classification are based on the relation set used in KEGG (Kyoto Encyclopedia of Genes and Genomes), so that pathway maps using KEGG notation can be automatically generated. We present the current version of the relation extraction algorithm and an evaluation based on a corpus of abstracts obtained from PubMed. The results indicate that the method is able to combine a reasonable coverage with high accuracy. We found that 61% of all sentences were parsed, and 97% of the parse trees were judged to be correct. The extraction algorithm was tested on a sample of 300 parse trees and was found to produce correct extractions in 90.5% of the cases.