-
Java 1.8.0_172 (Java 10 is not supported!)
-
dkpro-tc-0.9.0 Framework
/src/main/java/de.ba.tiagosenc.Pipeline/SdPipeline.java: pipeline to execute the program
SdPipeline line 58: to set your own DKPRO_HOME path before run the program. After execute the program all the information is gathered on the folder DKPRO_HOME.
SdPipeline line 63 to evaluate the system on cross validation
SdPipeline line 64: to evaluate the system on train/test
SdPipeline between Line 95 and 112: to select ontology-features and n-gram feature
SdPipeline line 92: to select classifier Support Vector Machine
/src/main/java/de.ba.tiagosenc.Features/: classes for the Ontology-features
/src/main/java/de.ba.tiagosenc.WikiParser/: classes for Jsoup Wikipedia anchor text extraction
/de.ba.tiagosenc/src/main/resources/Data/: Training/Test/Label tweets-data plus Spanish stopwords file.
/de.ba.tiagosenc/src/main/resources/Expanded Lists/: Information is directly extracted here, when using the “WikiParser” classes.
/de.ba.tiagosenc/src/main/resources/Lists/: Information is directly extracted here, when using the ontology-features.
/de.ba.tiagosenc/src/main/resources/Lists+Wiki/: contain the lists from “Expanded Lists” and “Lists”.
/de.ba.tiagosenc/src/main/resources/Lists+Wiki_Intro/: contain the lists from “Expanded Lists” and “Lists”, where the “Expanded Lists” contain just the anchor texts from the introduction of each Wikipedia article.
/src/test/java/: contain Junit-tests
/src/test/resources/: contain small content from the data files to test the reader