Abstract | ||
---|---|---|
We investigate modeling strategies for English code-switched words as found in a Swahili spoken term detection system. Code switching, where speakers switch language in a conversation, occurs frequently in multilingual environments, and typically de- teriorates STD performance. Analysis is performed in the context of the IARPA Babel program which focuses on rapid STD system development for under-resourced languages. Our results show that approaches that specifically target the modeling of code-switched words, significantly improve the detection performance of these words. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1016/j.procs.2016.04.040 | Procedia Computer Science |
Keywords | Field | DocType |
Spoken term detection,code switching,Swahili,pronunciation modeling | Pronunciation,Conversation,Computer science,Code-switching,Swahili,Speech recognition,Artificial intelligence,Natural language processing,System development | Conference |
Volume | ISSN | Citations |
81 | 1877-0509 | 0 |
PageRank | References | Authors |
0.34 | 8 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Neil Kleynhans | 1 | 15 | 1.52 |
William Hartmann | 2 | 64 | 10.66 |
Daniel R. van Niekerk | 3 | 18 | 4.69 |
Charl Johannes van Heerden | 4 | 133 | 12.50 |
Richard M. Schwartz | 5 | 2839 | 717.76 |
Stavros Tsakalidis | 6 | 213 | 13.83 |
Marelie H. Davel | 7 | 0 | 0.34 |