Fig 1.
Harmony search algorithm using dependency types and window of words for WSD.
Fig 2.
An example of generating typed dependencies.
“The dog scratched its back on the bark of the tree”.
Fig 3.
The collapsed form of the dependecy parses.
“The dog scratched its back on the bark of the tree”
Fig 4.
Harmony memory initialisation.
The pseudo code of initialising the harmony memory of HSA for WSD.
Fig 5.
The main steps of finding the best senses combination for an instance of WSD problem using HSA.
Table 1.
Comparison between the dependency types and window of words based on recall metric.
Table 2.
The results obtained from combining the typed dependencies with window of words.
Fig 6.
Context selection effectiveness.
HSA efficiency using various context selection methods over the selected set of files from SemCor 3.0.
Table 3.
Comparison of HSA to related works based on noun part-of-speech over nineteen files from SemCor corpus.
Table 4.
Comparison of HSA to related works based on the Senseval-2 dataset.