Dictionary Browser Cut Board Info Help English | Hungarian

The Dictionary Browser Page

This page introduces the features available on the Dictionary Browser Page. The description is divided into four sections according to the numbered features below.

1. The Dictionary Browser Page

1. List of the translation candidates ranked by their translation probabilities. This guarantees that most often used translations come first in the list (from top to bottom). Absolute corpus frequencies are also displayed.
   Click on one of the translation candidates to display the contexts it appears in (see [4]).
   
   
2. A plot displaying the distribution of the possible translations of the source word according to translation probability and the ratio of corpus ferquency between the source word and the corresponding translation candidate. 
   This kind of representation serves primarily testing purposes and gives hints about the "appropriateness of a translation". Presuming that translational synonyms show up exactly in the same contexts, in the case of translational synonymy the absolute corpus frequencies of the source and of the target lemmata should be approximately the same. Thus, those translation candidates can be treated as translational synonyms that appear with relatively high translation probability near 0 on the x-axis. Generally speaking, mostly nouns appear in the relevant range, especially nouns with definite denotata or nouns referring to natural types (e.g. augalas - LIT növény - HUN; 'plant').
   
   
3. Coloured target language word cloud reflecting semantic relations between source and target lemmata.
   Words in the word cloud vary in two ways:
   
   • First, their size depends on their translation probabilities: the higher the probability of the target word the bigger the font size is.
   • Secondly, colours are assigned to target words according to their frequency ratios relative to the source word: less frequent target words are cool-coloured (dark blue and light blue) while more frequent target words are warm-coloured (red, orange). Target lemmata of the same corpus frequency as the source word are displayed in grey. According to our hypothesis the frequency ratios provide hints about the semantic relation between source and target words. For instance, in Figure 1 Area 3 displays four Hungarian eqivalents of the Lithuanian lemma karieta. These are the following: kocsi (word with general meaning, e.g. 'car', 'railway wagon', 'horse-drown vehicle'), hintó ('carriage'), konflis ('a horse-drawn vehicle for public hire'), jármű ('vehicle'). The various colours of the candidates indicate different semantic relations. For instance, the red colour of kocsi marks that the meaning of the target word is more general than that of the source word. Conversely, the dark blue colour of konflis shows that the meaning of the target word is more special.
   
   A list of relevant translation candidates can be displayed in Area 1 by clicking any word in the word cloud. That is, the reversed dictionary can be queried by one click: the target lemma becomes the source lemma. In our recent example, translation candidates of kocsi can be displayed in Area 1 by clicking this word in the word cloud. As a result we get the following Lithuanian translation candidates: automobilis ('car'), mašina ('car'), vežimas ('horse-drawn vehicle'). And now, just as in the first case, relevant example sentences appear if we select any of the translation candidates (see [1]).
   
   
4. Provided example sentences with the source and target words highlighted. The examples can be displayed by clicking any of the translation candidates in [1]. This kind of information may be particularly useful for someone who wants to make utterances in a foreign language.