A study on detecting misleading online news using bigram and cosine similarity
Fake news can impact negatively in terms of creating negative perception towards business, organization, and government. One of the ways that fake news is created is through deceptive news writing. Many researchers have developed approaches in detecting deceptive news conten...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Science Publishing Corporation
2018
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/72998/ http://psasir.upm.edu.my/id/eprint/72998/1/FAKE.pdf |
| Summary: | Fake news can impact negatively in terms of creating negative perception towards business, organization, and government. One of the ways that fake news is created is through deceptive news writing. Many researchers have developed approaches in detecting deceptive news content using machine-learning approach and each of the approach has its own focus.Previous researches emphasis on the components of the news content such as in detecting grammar, humor, punctuation, body-dependent and body-independent features.In this paper, a new approach in detecting deceptive news based on misleading news has been developed which is focusing on the similarity between the content and its headlines using bigram and cosine similarity. Based on the experiments, the proposed approach has better performance in terms of detecting deceptive news. |
|---|