1887
Pattern-based Approaches to Semantic Relation Extraction
  • ISSN 0929-9971
  • E-ISSN: 1569-9994
USD
Buy:$35.00 + Taxes

Abstract

In the widely bi- and multilingual context of work in terminology and terminography today, and with the increasing volume of text-based resources available for carrying out this work, there is a need for computer tools to assist terminologists and terminographers in their tasks — including knowledge acquisition, conceptual description, creation of concept systems, formulation of definitions, and establishment of equivalence between terms — in two or more languages. Conceptual relations may be very useful for all of these applications. One method of semi-automatically locating information about conceptual relations uses knowledge patterns, i.e., linguistic markers of relations that can be used to find segments of text that convey this information. However, this kind of approach faces a number of serious challenges, including that of evaluating the certainty and thus ultimate usefulness of the information identified. Moreover, to date little research has been carried out to evaluate whether knowledge-pattern-based approaches may be expected to encounter comparable phenomena with similar frequency in different languages. This article will explore expressions of uncertainty observed in occurrences of English and French knowledge patterns identifying conceptual relations of ASSOCIATION and CAUSE–EFFECT, and how these may differ in the two languages.
Loading

Article metrics loading...

/content/journals/10.1075/term.14.1.07mar
2008-01-01
2019-09-19
Loading full text...

Full text loading...

References

http://instance.metastore.ingenta.com/content/journals/10.1075/term.14.1.07mar
Loading
  • Article Type: Research Article
Keyword(s): conceptual relations , English , French , hedging , knowledge patterns , modal verbs , negation and quantification
This is a required field
Please enter a valid email address
Approval was successful
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error