Adapted Weighted Graph for Word Sense Disambiguation

In Natural Language Processing, Word Sense Disambiguation is defined as the task to assign a suitable sense of words in a certain context. Word Sense Disambiguation takes an important role and considered as the core research problem in computational linguistics. In this research, we conduct an experiment with Adapted Lesk Algorithm compared to original Lesk Algorithm to improve the performance of weighted graph-based word sense disambiguation. Both Algorithms base their measure to the gloss of the dictionary used, not like the other similarity measure that base their measure to the path or information content of the concept being compared. Thus, both Lesk and Adapted Lesk has the highest coverage of part-of speech since they can measure between different part-of-speech. Results of the experiment indicate that Adapted Lesk improves the performance of weighted graph-based Word Sense Disambiguation by 19 % of precision compared to Original Lesk in individual similarity measure experiment.