NUK - logo
Faculty of Arts, Lj. - all departments (FFLJ)
OHK FF oddelčne knjižnice bodo imele v času poletnih mesecev nekoliko prilagojene urnike. Poletne urnike si lahko ogledate:
Poletni urniki
  • Semantično označevanje korpusov
    Fišer, Darja, 1978-
    Semantic annotation of corpora is the process of assigning meanings to words in a corpus by taking into account the context in which they appear. Semantically annotated corpora are indispensible in ... natural language processing tasks, such as automatic word sense disambiguation, information retrieval and machine translation. In addition, they are also extremely useful in applied linguistics tasks, such as lexicography and language pedagogy, as well as in corpus linguistics for the study of sense frequency and co-occurrence. However, semantic annotation is hard, slow and expensive; in many cases it is difficult to pin down the meaning of a word or draw the boundaries between two similar meanings, and it is even less clear how specific sense assignment should be. This is why only a few semantically annotated corpora are currently available for English and very few other languages. For Slovene, no previous attempt has been made to obtain such a corpus. This paper presents and discusses a project in which the most frequent nouns from a corpus of Slovene were manually annotated with wordnet senses. The evaluation of the annotation shows that wordnet senses are often to fine-grained for reliable sense assignment, which is why we present a technique to find the most similar senses and merge them into larger sense categories that simplify the annotation process as well as improve the inter-annotator agreement.
    Source: Slovenske korpusne raziskave (Str. 110-130)
    Type of material - article, component part ; adult, serious
    Publish date - 2010
    Language - slovenian
    COBISS.SI-ID - 43099234