Mohammad R. Eletriby, T. Reynolds, Ramesh C. Jain, Kai Zheng
{"title":"研究在线消费者健康文本中上下文信息的命名实体识别","authors":"Mohammad R. Eletriby, T. Reynolds, Ramesh C. Jain, Kai Zheng","doi":"10.1109/INTELCIS.2017.8260078","DOIUrl":null,"url":null,"abstract":"Online health forums have become a common place for healthcare consumers (e.g., patients, caregivers) to seek information and exchange support. One challenge is that patients may not know how to frame a question properly, especially on what contextual information to disclose that would help the online community better understand the health concerns they have. In this study, we analyzed the contextual information disclosed by users of the Lung and Respiratory Disorders community in a popular online health forum, Medhelp.org. We analyzed both questions and answers to understand what contextual information tends to be often missing from the questions that may hinder communication effectiveness. In doing so, we also compared two different natural language processing approaches: (1) MetaMap developed by the U.S. National Library of Medicine, and (2) IBM Natural Language Classifier (NLC), to examine their respective performance when applied to consumer health text. Our results show that the two methods are complementary, and combining them together would result in a high-performing recognition tool with an overall F-score of 80.4%.","PeriodicalId":321315,"journal":{"name":"2017 Eighth International Conference on Intelligent Computing and Information Systems (ICICIS)","volume":"99 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":"{\"title\":\"Investigating named entity recognition of contextual information in online consumer health text\",\"authors\":\"Mohammad R. Eletriby, T. Reynolds, Ramesh C. Jain, Kai Zheng\",\"doi\":\"10.1109/INTELCIS.2017.8260078\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Online health forums have become a common place for healthcare consumers (e.g., patients, caregivers) to seek information and exchange support. One challenge is that patients may not know how to frame a question properly, especially on what contextual information to disclose that would help the online community better understand the health concerns they have. In this study, we analyzed the contextual information disclosed by users of the Lung and Respiratory Disorders community in a popular online health forum, Medhelp.org. We analyzed both questions and answers to understand what contextual information tends to be often missing from the questions that may hinder communication effectiveness. In doing so, we also compared two different natural language processing approaches: (1) MetaMap developed by the U.S. National Library of Medicine, and (2) IBM Natural Language Classifier (NLC), to examine their respective performance when applied to consumer health text. Our results show that the two methods are complementary, and combining them together would result in a high-performing recognition tool with an overall F-score of 80.4%.\",\"PeriodicalId\":321315,\"journal\":{\"name\":\"2017 Eighth International Conference on Intelligent Computing and Information Systems (ICICIS)\",\"volume\":\"99 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2017-12-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"3\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2017 Eighth International Conference on Intelligent Computing and Information Systems (ICICIS)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/INTELCIS.2017.8260078\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 Eighth International Conference on Intelligent Computing and Information Systems (ICICIS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/INTELCIS.2017.8260078","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Investigating named entity recognition of contextual information in online consumer health text
Online health forums have become a common place for healthcare consumers (e.g., patients, caregivers) to seek information and exchange support. One challenge is that patients may not know how to frame a question properly, especially on what contextual information to disclose that would help the online community better understand the health concerns they have. In this study, we analyzed the contextual information disclosed by users of the Lung and Respiratory Disorders community in a popular online health forum, Medhelp.org. We analyzed both questions and answers to understand what contextual information tends to be often missing from the questions that may hinder communication effectiveness. In doing so, we also compared two different natural language processing approaches: (1) MetaMap developed by the U.S. National Library of Medicine, and (2) IBM Natural Language Classifier (NLC), to examine their respective performance when applied to consumer health text. Our results show that the two methods are complementary, and combining them together would result in a high-performing recognition tool with an overall F-score of 80.4%.