Java Reference
In-Depth Information
Summary
We have discussed the parsing process and how it can be used to extract relationships from
text. It can be used for a number of purposes including grammar checking and machine
translation of text. There are numerous text relations possible. These include such relation-
ships as father of, near to, and under. They are concerned with how elements of text are re-
lated to each other.
Parsing the text will return relationships that exist within the text. These relationships can
be used to extract information of interest. We demonstrated a number of techniques using
the OpenNLP and Stanford APIs to parse text.
We also explained how the Stanford API can be used to find coreference resolutions within
text. This occurs when two or more expressions, such as "he" or "they", refer to the same
person.
We concluded with an example of how a parser is used to extract relations from a sentence.
These relations were used to extract information to answer simple "who" type queries about
U.S. presidents.
In the next chapter, we will investigate how the techniques developed in this and the previ-
ous chapters can be used to solve more complicated problems.
