A Cognitive Approach to Word Sense Disambiguation
An unsupervised, knowledge-based, parametric approach to Word Sense Disambiguation is proposed based on the well-known cognitive architecture ACT-R. In this work, the target word is disambiguated based on surrounding context words using an accumulator model of memory search and it is realized by incorporating RACE/A with ACT-R 6.0. In this process, a spreading activation network is built following the strategies of Tsatsaronis et al. proposed in  using the chunks and their relations in the declarative memory system of ACT-R and the lexical representation has been achieved by integrating WordNet with the cognitive architecture. The resulting Word Sense Disambiguation system is evaluated using the test data set from English Lexical Sample task of Senseval-2 and overall accuracy of the proposed algorithm is 44.74% which outperforms all the participating Word Sense Disambiguation Systems.
KeywordsWSD Word Sense Disambiguation RACE/A Retrieval by ACcumulating Evidence in an Architecture ACT-R Adaptive Control of Thought-Rational SAN Spreading Activation Network
Unable to display preview. Download preview PDF.