Line 18: | Line 18: | ||
== Abstract == | == Abstract == | ||
− | + | Text simplification is a field that can be useful for many applications because it can improve text readability and help in NLP tasks. Thus, I will give a brief overview about this field. Then, I will talk about the specific problem that we are trying to solve, which is the creation of a text simplification system with educational purposes. We have the simplification of Geography and History texts of 7 th to 8th written in Portuguese to 3 th and 4 th graders as initial goal. Given the complexity of this problem, we decided to break it into subtasks. Hence, we are exploring ways of performing the most frequent syntactic simplification: splitting sentences. I will also discuss about the initial work, such as the creation of simplified corpora, adaptation/creation of splitting sentence classifiers and I will compare with similar work developed for English and Brazilian Portuguese. | |
− | + | ||
'''Note:''' This seminar will be held in English, if required. | '''Note:''' This seminar will be held in English, if required. |
Text simplification is a field that can be useful for many applications because it can improve text readability and help in NLP tasks. Thus, I will give a brief overview about this field. Then, I will talk about the specific problem that we are trying to solve, which is the creation of a text simplification system with educational purposes. We have the simplification of Geography and History texts of 7 th to 8th written in Portuguese to 3 th and 4 th graders as initial goal. Given the complexity of this problem, we decided to break it into subtasks. Hence, we are exploring ways of performing the most frequent syntactic simplification: splitting sentences. I will also discuss about the initial work, such as the creation of simplified corpora, adaptation/creation of splitting sentence classifiers and I will compare with similar work developed for English and Brazilian Portuguese.
Note: This seminar will be held in English, if required.