Research in Computing Science, Vol. 70, pp. 223-232, 2013.
Abstract: In the past few years, there has been a growing interest in detecting quotation families and in automatically analyzing transformation patterns. However, no work has provided a complete qualitative analysis of these variations. Through a comprehensive linguistic analysis, the goal of this paper is to study and categorize the way quotes from newspapers tend to evolve and deform over time. In order to help in observing patterns and variability, we apply global sequence alignment techniques, commonly used in bioinformatics. Recurrent patterns, such as the common deletion of words expressing modality, paraphrases, or strong synonymic variations, are listed and discussed. In addition to providing a better understanding of cultural dynamics in media, we believe that the categorization of variation regularities in quotes can help further enhance the quality of similar quotations clustering algorithms and other NLP tasks such as paraphrase identification.
Keywords: Discourse analysis, quotes, memes, multiple sequence alignment, linguistics, co-reference, modality
PDF: Analyzing Variation Patterns In Quotes Over Time
PDF: Analyzing Variation Patterns In Quotes Over Time