The problem is formulated in terms of obtaining the minimum description length of a text, and the proposed solution finds the segments and their languages through dynamic programming .

An actual procedure for obtaining an optimal result through dynamic programming was proposed.

We can straightforwardly implement this recur—sive computation through dynamic programming , by managing a table of size |X | x To fill a cell of this table, formula (4) suggests referring to t x |£| cells and calculating the description length of the rest of the text for O( |X | —t) cells for each language.

Nevertheless, since we use a uniform amount of training data for every language, and since varying 7 would prevent us from improving the efficiency of dynamic programming , as explained in §4, in this article we set 7 to a constant obtained empirically.

By applying the above methods, we propose a solution to formula (1) through dynamic programming .

In the figure, phone bigram TF—IDF is labeled p2; phonetic alignment with dynamic programming is labeled DP.

The feature selection experiments in Figure 2 shows that the TF—IDF features alone are quite weak, while the dynamic programming alignment features alone are quite good.

Given (13,10), we use dynamic programming to align the surface form 1‘9 with all of the baseforms of w. Following (Riley et al., 1999), we encode a phoneme/phone with a 4-tuple: consonant manner, consonant place, vowel manner, and vowel place.