A Network-Based Spam Detection Framework for Reviews in Online Social Media
K. Amar1, M. Kameshwara Rao2, Ch. Chaitanya3, Ravi Kumar Tenali4
1K. Amar, Department of ECM, Koneru Lakshmaiah Education Foundation, Vaddeswaram (A.P), India.
2Ch. Chaitanya, Department is ECM, Koneru Lakshmaiah Education Foundation, Vaddeswaram (A.P), India.
3Ravi Kumar Tenali, Department is ECM, Koneru Lakshmaiah Education Foundation, Vaddeswaram (A.P), India.
4Dr. M.Kameswara Rao, Department is ECM, Koneru Lakshmaiah Education Foundation, Vaddeswaram (A.P), India.
Manuscript received on 07 April 2019 | Revised Manuscript received on 20 April 2019 | Manuscript published on 30 April 2019 | PP: 748-752 | Volume-8 Issue-6, April 2019 | Retrieval Number: F3645048619/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Now-a-days, individuals mostly depend on the content in social media in decision making. For instance, they choose to purchase an item depending on the reviews and feedback. Possibility of leaving a review gives a golden chance for spammers to put in writing spam reviews concerning the product and services for various interests. Distinguishing these spammers and accordingly the spam content could be an interesting issue of analysis. Though a substantial range of studies are done recently towards this, till date the methodologies used still barely find the spam reviews. Here, we propose a unique framework called Net-Spam that utilizes spam options for modelling review datasets as heterogeneous data networks to map spam detection procedure into classification. Mistreatment the importance of spam options helps us to get best output in terms of various metrics experimented on real-world review datasets from Yelp and Amazon websites. The results show that Net-Spam outperforms the prevailing ways and among four classes of features; including review-behavioural, user-behavioural, review linguistic, user-linguistic, the primary options performs better than the other categories.
Keyword: Social Media, Spammers, Review, Framework, Net-Spam, Heterogeneous data Networks.
Scope of the Article: Network Based Applications