Parallel Text Alignment
An interesting problem in media synchronization is the alignment of parallel texts, multiple translations of the same document. Parallel texts are differing presentations of the same content. An alignment among parallel texts is a mapping of content in one text to content in another wherein the subjects of the mapping are translations of the same range in the original text. Discovery of the alignment between the texts allows comparative analysis of translation style and semantics as well as retrieval of content from multiple translations based on a search of an original document or an alternative translation. Two versions of this problem exist: aligning cross-language or aligning infra-language. Cross-language alignment is the alignment of two documents which are not in the same language, Spanish to English for example. Intra-language alignment is the alignment of two documents in the same language. This chapter specifically addresses the problem of computed infra-language alignment, though a discussion of cross-language alignment is included.
KeywordsMultimedia Application Rubber Band Text Segment Parallel Text Word Score
Unable to display preview. Download preview PDF.