taco is a toolkit for implementing TArget COnstraints in a target-constrained statistical machine translation (SMT) system, such as that described in
"Agreement Constraints for Statistical Machine Translation into German" Philip Williams and Philipp Koehn, in WMT 2011.
and
"Using Feature Structures to Improve Verb Translation in English-to-German Statistical MT" Philip Williams and Philipp Koehn, in HyTra 2014.
The toolkit provides a language-independent library, libtaco, intended for use with a decoder that supports string-to-tree or tree-to-tree translation (such as Moses). It also provides a number of language- and annotation-specific tools for extracting target constraints and feature structure lexicons from training data.
...
Boost 1.48 or later.
$ autoreconf -i $ ./configure --with-boost=/path/to/boost --prefix=/path/to/dest $ make && make install