BDPA

In this section we offer a collection of manually edited pairwise alignments. For details on the dataset, see the descriptions in the table below.

Name	Size	Description
Covington	82 alignments	Dataset by Covington (1996), here in a slightly modified form containing plain IPA transcriptions instead of the original hybrid format.
GLOBAL	7197 alignments	Dataset was automatically extracted from our masterset of multiple alignments by selecting the most diverse sequence pairs of all multiple alignments. Parts of this dataset were used in order to test the performance of the SCA alignment algorithm (List 2012).
TONE	1088 alignments	Dataset was automatically extracted from our masterset by selecting the most diverse sequence pairs of all multiple alignments consisting of tone languages. In contrast to the GLOBAL dataset above, this dataset contains only phonetic sequences from tone languages (Chinese, Bai).