Distilling Task Knowledge from How-to Communities
You can find more information about HowToKB here.
Before Running Code
You need to install file javatools-1.0.0.jar (in jars folder) into your maven library, following this guide: https://maven.apache.org/guides/mini/guide-3rd-party-jars-local.htmlCommandline:
mvn install:install-file -Dfile=jars/javatools-1.0.0.jar -DgroupId=niket.tools -DartifactId=javatools -Dversion=1.0.0 -Dpackaging=jar
How To Run
- Download HowToKB dataset from here and put them into howtokb-data folder
- Moved the w2v model file (articles-word2vec-word-pos.model.txt) to "data" folder
- Set "data" and "resources" folders as source folders in the configuration (build path -> use as source folders)
- Some testing classes:
- test.kb.howtokb.TaskFrameReaderTest: reading task frames from howtokb data
- test.kb.howtokb.WikiHowArticleReaderTest: reading wikihow articles from howtokb data
- test.kb.howtokb.TextToWikiHowTaskFrameTest: extracting task frames from wikihow articles
- test.kb.howtokb.jdbc.JDBCConnectorTest: query data from howtokb database
- test.kb.howtokb.clustering.*: clustering task frames