#Precision: An Exploration of the Utility of User-Generated Metadata for the Creation of Precise Microblog Query-Expansion Systems

Twitter research provides a unique opportunity to answer fundamental questions regarding the best methods for the large-scale retrieval of extremely sparse documents. This study examines the utility of user-generated metadata expansion candidate terms for the creation of precise microblog search engines. Several search engines were created utilizing different genres of candidate expansion terms, confidence thresholds, and document parameters to explore this issue. This study demonstrates that user-generated metadata has utility for the precise retrieval of terse queries with high levels of associated conversation, such as movie awards or current events, but performs poorly on textually rich queries with lower levels of perceived conversation.Master of Science in Information Scienc