The project to be described integrates “linguistic registers” in retrieval research to explore the impact on retrieval set membership, ranking options, and presentation opportunities to optimize ranking results presented the end‐user.Ģ. PREVIOUS WORK ‐ THEORIES OF REGISTER AND RETRIEVAL, COMPREHENSIBILITY AND REASONS FOR EXPLORATIONĬonsider for instance a corpus of medical terms derived from various literatures, representing scientific, practice, and consumer health. We start by describing the theory of register and warrant such an approach for IR. The ideas are explored in a project implementing some of the linguistic behaviors defined in ISO12620:2009 to determine whether LR‐IR is a sustainable research program. ISO12620:2009 proposes a “Terminology Classification Model” titled “Specification of data categories and management of a Data Category Registry for language resources” ( ). As an understudied area we feel it is important first to identify briefly parallels between IR practice and LR‐IR and next to suggest IR framework models that would be sensitive to a semantic‐level linguistic‐oriented one. But an ISO standard on “ling reg” seems appropriate also for IR. The usual interpretation of linguistic register seems to be merely identifying parts of speech as a natural language project or gross‐level syntax parsing, shifting IR into computational linguistics. Looking for ways to improve the comprehensibility and appropriateness of resource location, we consider a “linguistic register” (LR‐IR, or “ling‐reg”) approach, which turns out to be not without some controversy. Recent efforts (Spring 2020) to keep the public and health care practitioners informed with covid‐19 data that lead to healthier action in their daily lives have led to a confusing miasma of contradictory facts and mix of research reports, guidelines, blogs and editorials. We conclude that developing linguistic register‐oriented IR and visualization is a promising approach to a new IR model. Applying visualization techniques, we see the impact of linguistic register on distribution by domain and purpose. The results suggest strongly that applying ISO12699:2019 as part of a retrieval model reveals unanticipated concepts in the collection and exposes fluidity in term‐use, as end‐user linguistic performance. In this paper we suggest an interdisciplinary model for retrieval engine design that is sensitive to the seekers' own linguistic registers and information needs, using covid‐19 as the example. Certainly the rapid developments of covid‐19 provide challenge and opportunity for evaluating different techniques for public awareness. The challenges of producing data resources, sharing them, and sifting actionable data from presidential “sarcasm” raise novel challenges.
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. Archives
December 2022
Categories |