Automated Detection of Adverse Drug Reactions from Social Media Posts with Machine Learning

被引：18

作者：

Alimova, Ilseyar ^{[1
]}

Tutubalina, Elena ^{[1
]}

机构：

[1] Kazan Volga Reg Fed Univ, Lab Chemoinformat & Mol Modeling, Kazan, Russia

来源：

ANALYSIS OF IMAGES, SOCIAL NETWORKS AND TEXTS, AIST 2017 | 2018年 / 10716卷

基金：

俄罗斯科学基金会;

关键词：

Adverse drug reactions; Text mining Health social media analytics; Machine learning; Deep learning; HOSPITALIZED-PATIENTS; EVENTS; CORPUS; PHARMACOVIGILANCE;

D O I：

10.1007/978-3-319-73013-4_1

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Adverse drug reactions can have serious consequences for patients. Social media is a source of information useful for detecting previously unknown side effects from a drug since users publish valuable information about various aspects of their lives, including health care. Therefore, detection of adverse drug reactions from social media becomes one of the actual tools for pharmacovigilance. In this paper, we focus on identification of adverse drug reactions from user reviews and formulate this problem as a binary classification task. We developed a machine learning classifier with a set of features for resolving this problem. Our feature-rich classifier achieves significant improvements on a benchmark dataset over baseline approaches and convolutional neural networks.

引用

页码：3 / 15

页数：13