This paper addresses the challenge of detecting medical misinformation online by presenting a new labeled dataset of misinformative and non-misinformative comments from a health discussion forum. The dataset was created through information retrieval techniques and a detailed labeling strategy. By employing Recursive Feature Elimination to identify the most relevant features, the study achieved a classification accuracy of 90.1%. This dataset, with a high proportion of non-misinformative comments (85.8%), aims to support the development of machine learning systems for identifying misinformation in user-generated medical content.
-
Kinsora, Alexander, et al. "Creating a labeled dataset for medical misinformation in health forums." 2017 IEEE International Conference on Healthcare Informatics (ICHI). IEEE, 2017.
-
Members
- Madhouse
- JoelR
- Steph40
- eivindsimensen
- Matt
- bernhara
- PrettyPixels
- Sinistra
- onlyME
- TomCat
- V0RT3X
- Square Wheels
- IC Essentials
- master963
- We are Borg
- 666wicked666
- Jon Erickson
- envy
- MissB
- aLEX49566
- TheJimmo
- Patreon Lukazuki
- pat
- burnyourfeelings
- JoeyM
- Labi
- Mitsuru
- A Zayed
- Adriano Faria
- Kirill Gromov
- Drufuss
- Charlie Feigel
- opentype
- kmk
- ali hagi
- Hong98
- ijinxcxx
- ijinxcxx4k
- MichaelR
- PPlanet
- ButterflyPixel
- Luki
- Labis
- Voyage
- flrn
- rivi235
- Ryan
- StevenM
- Aleksandar Markovic
- DawPi
- Como
- BEASTBOOSTER
- GazzaGarratt
- Roblox County DOJ Roleplay
- HDiddy
- Analog
- adik
- Dilip
- Destructor
- AnonDoggo
- terabyte
- Dani Onvlee
- Anthony Feng
- isvans
- MythonPonty
- GrantHorizons
- abobader
- eliteone
- ArashDev
- Brian
- Cory McElroy
- Richard Arch
- ReyDev
- Videoflicks
- Empire
- Nebulous
- Split
- aXenDev
- Astronis
- ITV
- Denis Dyack
- Claudia999
- Ticaga
- npnchicago
- Chris Anderson
- NewVicious
- UrbanNest Realtors
- Ioannis D
- Paul Cojocariu
- Maxius
- Yurii
- rnorth6920
- Dancho Danchev
- Voidcraft Studios
- Live Games
- Bob Schwarz
- bdmusic 24
- AdamOnTech
- DaninMS
- Ticulica