Generating Query Suggestions for Cross-language and Cross-terminology Health Information Retrieval
Autor: | Paulo Miguel Santos, Carla Teixeira Lopes |
---|---|
Rok vydání: | 2020 |
Předmět: |
Vocabulary
Information retrieval Computer science media_common.quotation_subject Unified Medical Language System 02 engineering and technology language.human_language Expression (mathematics) Terminology 020204 information systems 0202 electrical engineering electronic engineering information engineering language 020201 artificial intelligence & image processing Relevance (information retrieval) Portuguese Baseline (configuration management) Cross-language information retrieval media_common |
Zdroj: | Lecture Notes in Computer Science ISBN: 9783030454418 ECIR (2) |
Popis: | Medico-scientific concepts are not easily understood by laypeople that frequently use lay synonyms. For this reason, strategies that help users formulate health queries are essential. Health Suggestions is an existing extension for Google Chrome that provides suggestions in lay and medico-scientific terminologies, both in English and Portuguese. This work proposes, evaluates, and compares further strategies for generating suggestions based on the initial consumer query, using multi-concept recognition and the Unified Medical Language System (UMLS). The evaluation was done with an English and a Portuguese test collection, considering as baseline the suggestions initially provided by Health Suggestions. Given the importance of understandability, we used measures that combine relevance and understandability, namely, uRBP and uRBPgr. Our best method merges the Consumer Health Vocabulary (CHV)-preferred expression for each concept identified in the initial query for lay suggestions and the UMLS-preferred expressions for medico-scientific suggestions. Multi-concept recognition was critical for this improvement. |
Databáze: | OpenAIRE |
Externí odkaz: |