Abstract-The sharing of information has been proven to be beneficial for business partnerships in many application areas such as business planning or marketing. Today, association rule mining imposes threats to data sharing, since it may disclose patterns and various kinds of sensitive knowledge that are difficult to find. Such information must be protected against unauthorized access. The challenge is to protect actionable knowledge for strategic decisions, but at the same time not to lose the great benefit of association rule mining. To address this challenge, a sanitizing process transforms the source database into a released database in which the counterpart cannot extract sensitive rules from it. Unlike existing works that focused on hiding sensitive association rules at a single concept level, this paper emphasizes on building a sanitizing algorithm for hiding association rules at multiple concept levels. Employing multi-level association rule mining may lead to the discovery of more specific and concrete knowledge from datasets. The proposed system uses genetic algorithm as a biogeography-based optimization strategy for modifying multi-level items in database in order to minimize sanitization's side effects such as non-sensitive rules falsely hidden and fake rules falsely generated. The new approach is empirically tested and compared with other sanitizing algorithms depicting considerable improvement in completely hiding any given multi-level rule that in turn can fully support security of database and keeping the utility and certainty of mined multi-level rules at highest level.Index Terms-Database sanitization, genetic algorithm, privacy preserving data mining, multi-level association rule hiding.

No abstract

No abstract

The rapidly increasing of sentiment analysis in social networks has lead business owners and decision makers to value opinion leaders who can influence people's impressions concerning certain business or commodity. Nevertheless, decision makers are being misled by inaccurate results due to the ignorance of perspectivism. Considering perspectivism, while computing text polarity, can help machines to reflect the human perceived sentiment within the content. This emphasises the need for integrating social behaviour (user's influence factor) with sentiment analysis (text polarity scores), providing a more pragmatic portrayal of how the writer's audience comprehend the message. In this study, a new model is proposed to intensify sentiment analysis process on Twitter. In the achievement of such, social network analysis is done using UCINET tool followed by artificial neural networks for ranking users. For sentiment classification, a hybrid approach is presented, where lexicon-based technique is combined with a fuzzy classification technique to handle language vagueness as well as for an inclusive analysis of tweets into seven classes; for the purpose of enhancing final results. The proposed model is practiced on data collected from Twitter. Results show a significant enhancement in tweets polarity scores represent more realistic sentiments.