Simplifying Text in Medical Literature
Ethel Ong | Jerwin Damay | Gerard Lojico | Kimberly Lu | Dex Tarantan
A text simplification system is developed that accepts as input a document written for college reading level, and converts it to high school reading level text using techniques in natural language processing. The achieved level of lexical and syntactic simplifications is dependent on the lexical and syntactic information stored in the system's knowledge sources. These knowledge sources can be updated when necessary, thus allowing the system to support varying input and output reading levels. Test results show that simplification can increase the reading ease score from as low 2.5% to a high 296%. Results also show that only 70% of the generated easier-to-read text will have a reading ease score higher than the original input sentence.