The objective of this project is the construction of alternative methods for the Algorithm of Context commonly used to estimate the context function for Variable Length Markov Chains. We will apply these alternative methods for modern texts of Brazilian and European Portuguese authors. These texts had been codified according to whether its syllables are beginnings of prosodic words and if they were stressed or not. Applying the Context Algorithm to these texts we verify that the sample was not large enough to discriminate between the languages. Alternative methods of estimation are necessary.
News published in Agência FAPESP Newsletter about the scholarship: