S3 Dataset

S3 Dataset with Generation Code and Resources

Contents

1. Datasets
2. MedlinePlus 1. MedlinePlus-SemRep.xls
   • A two-column table (in .xls format) that presents the annotated version of each sentence in the MedlinePlus dataset along with the proposed simplification. 2. MedlinePlus-SemRep.xml
   • For each sentence in the MedlinePlus dataset, we provide:
     1. Its annotated version
     2. Its simplification
     3. The triples that have been identified in the original sentence
3. WikiAstronauts 1. WikiAstronauts-DBpedia.xls
   • A two-column table (in .xls format) that presents the annotated version of each sentence in the WikiAstronauts dataset along with the proposed simplification. 2. WikiAstronauts-DBpedia.xml
   • For each sentence in the WikiAstronauts dataset, we provide:
     1. Its annotated version
     2. Its simplification
     3. The triples that have been identified in the original sentence
4. Evaluations
5. MedlinePlus-SemRep.xls
6. WikiAstronauts-DBpedia.xls * The tables (in .xls format) that have been used to evaluate the methods according to which a simplification is selected.
7. src
8. CrowdFlower 1. MedlinePlus
   1. experiment.csv
   • The table (in .csv format) that was used on the CrowdFlower platform.
   • It contains the annotated version of the sentences from the MedlinePlus dataset.
   2. f902529.csv
   • The table (in .csv format) that contains all the simplifications that have been submitted by the contributors. 2. WikiAstronauts
   1. experiment.csv
   • The table (in .csv format) that was used on the CrowdFlower platform.
   • It contains the annotated version of the sentences from the WikiAstronauts dataset.
   2. f900315.csv
   • The table (in .csv format) that contains all the simplifications that have been submitted by the contributors.
9. Data 1. MedlinePlus 2. WikiAstronauts
   • Folders that contain (in .csv and .xml format) the original sentences from the two datasets along with their respective annotated versions and the triples.
10. Dataset-MedlinePlus.py * Python script that parses the original sentences that are found at ./Data/MedlinePlus/XML. * Provides various statistics regarding the dataset. * Stores the CrowdFlower table (in .csv format) at ./CrowdFlower/MedlinePlus/experiment.csv.
11. Dataset-WikiAstronauts.py * Python script that parses the original sentences that are found at ./Data/WikiAstronauts/XML. * Provides various statistics regarding the dataset. * Stores the CrowdFlower table (in .csv format) at ./CrowdFlower/WikiAstronauts/experiment.csv.
12. Output-MedlinePlus.py * Python script that processes the simplifications that are proposed by the contributors and are located at ./CrowdFlower/MedlinePlus/f902529.csv. * Implements a variety of metrics in order to choose the most appropriate simplification. * By executing python Output-MedlinePlus.py --evaluation, it selects 30 random sentences that are used for the evaluation purposes and are stored by default at ../Evaluation/MedlinePlus-SemRep.xls.
13. Output-WikiAstronauts.py * Python script that processes the simplifications that are proposed by the contributors and are located at ./CrowdFlower/WikiAstronauts/f900315.csv. * Implements a variety of metrics in order to choose the most appropriate simplification. * By executing python Output-WikiAstronauts.py --evaluation, it selects 30 random sentences that are used for the evaluation purposes and are stored by default at ../Evaluation/WikiAstronauts-DBpedia.xls.

License

This project is licensed under the terms of the Apache 2.0 License.