The rise and fall of stopword lists in patent searching and mapping
Abstract
Stopword lists are collections of so-called 'noise' words, which appear frequently in documents but are considered as non information-bearing. They have been introduced as a way to enhance information retrieval e.g. patent information retrieval in 1958 and were further developed until the 1970's. Since then, they have remained almost unchanged albeit they have gained importance by forming the basis for advanced tools such as text mining, mapping or clustering tools. Moreover, they are currently mostly used in static form whereas they could benefit from the latest findings of information science.
Domains
Information Retrieval [cs.IR]
Origin : Files produced by the author(s)