Co-occurrence

Co-occurrence referred to in General Linguistics, the common occurrence of two lexical units (eg words) in a parent unit, as in a sentence or a document. There is the assumption that these two terms are mutually dependent if they strikingly often occur together. Statistical tests provide measures of the expected dependence, such as various variants of mutual information and likelihood-ratio tests. This can have both grammatical and semantic causes.

If a grammatical or semantic dependency of two frequently occurring terms exists together proven, it is called collocation.

Both concepts have great importance in information retrieval.

Examples of co-occurrences

1 I'm sitting in the bank and I go to the bank and I'm sitting on the bench brings the verbs sit and not go randomly with the database object in a context. Only in the second example sentence is still an ambiguity. The ambiguity is not resolved by the verb, as it still can be a bench or a building. However, depends not sit on bench, it could also be a chair - but the chance to sit on a bench is more than sitting up on a line or to cook the bank.

2 sayings, however, are fixed Kookkurenzen because it is rigid expressions, such as: it raining cats and dogs.

3 The expected probability is very high when a when present in a set that is followed in order to express a causal relationship with precondition and conclusion. As you can see, the existential interconnected concepts do not stand behind each other, they are, however, logically in a sequence.

Set and Nachbarschaftskookkurrenz

In the practice of text mining, a distinction between Satzkookkurrenz ( lexical items occur together in a block ) and Nachbarschaftskookkurrenz ( lexical units are next to each other ). It could also mean viewing in larger text contexts ( paragraph or document co-occurrence ), in practice, these are not considered, not least because of the high computational complexity in machine processing.

Source

485521
de