The project pipeline can be found in WRAP.sh. Briefly..
1. To extract corpora from the AcqDiv database: 1.split.clean.phonemize.R
2. To remove child utterances: 2.prepareAcqdiv.py
3. To segment: 3.segment.sh
4. To collapse segmentation results: 4.collapse.results.sh
5. To present results: 5.ACL.markdown.Rmd
-> Results in overunder_new.csv and stats.csv.