Spam Fighting in Social Tagging Systems

Abstract

Abstract. Tagging in online social networks is very popular these days, as it facilitates search and retrieval of diverse resources available online. However, noisy and spam annotations often make it difficult to perform an efficient search. Users may make mistakes in tagging and irrelevant tags and resources may be maliciously added for advertisement or self-promotion. Since filtering spam annotations and spammers is time-consuming if it is done manually, machine learning approaches can be employed to facilitate this process. In this paper, we propose and analyze a set of distinct features based on user behavior in tagging and tags popularity to distinguish between legitimate users and spammers. The effectiveness of the proposed features is demonstrated through a set of experiments on a dataset of social bookmarks.

Securing Tag-based recommender systems against profile injection attacks: A comparative study. (Extended Report)

“…To filter out non-legitimate users, various anti-spamming techniques have been developed by numerous researchers in the field, employing either a form of Bayesian type filtering (Yazdani et al, 2012), or other tag classification methods . Particular characteristics of the tags used in annotations are exhibited, such as the presence of suspicious ones into user profiles.…”

“…In the above works the assumption made is that, any tags used by a legitimate user would coincide with those by other legitimate users. Nevertheless, if attackers are aware of the correlations being used in an attack filtering policy, they might try to disguise their fake profiles by using non-suspicious tags Yazdani et al, 2012). In such a case, any spam tag filtering is turned ineffective.…”

“…To reduce processing time, for every run we selected randomly subsets of folksonomies corresponding to 3.000 users. As opposed to other researchers who used pre-labeled datasets with spam data (e.g., Yazdani et al (2012); Poorgholami et al (2013); Zhai et al (2016)), in our case, due to the absence of such bogus data we chose to evaluate over synthetic ones, we generated out of the original. Moreover, this enabled us to study exclusive cases of attacks.…”

Quality-protected folksonomy maintenance approaches: a brief survey

“…Spammers usually use different techniques for polluting the folksonomy. Examples include, by creating many accounts, publishing a specific post many times, using semantically diverse tags to annotate a resource and teaming up with other spammers to give good rating to each other (Yazdani et al ., 2012).…”