The SBD process is language-dependent and is often not straightforward. Common approaches to detect sentences include using a set of rules or training a model to detect them. A set of simple rules for detecting a sentence follows. The end of a sentence is detected if the following is true:
- The text is terminated by a period, question mark, or exclamation mark
- The period is not preceded by an abbreviation or followed by a digit
Although this works well for most sentences, it will not work for all of them. For example, it is not always easy to determine what an abbreviation is, and sequences such as ellipses may be confused with periods.
Most search engines are not concerned with SBD. They are only interested in a query's tokens and their positions. POS-taggers and other NLP tasks that perform the extraction of data will frequently process individual sentences. The detection of sentence boundaries will help separate phrases that might appear to span sentences. For example, consider...