Conclusions

Abstract

We have presented a new algorithm for generating trust networks based on user tagging information to make recommendations in the online environment. The experiment results showed that this tag-based similarity approach performs better while making recommendations than the traditional collaborative filtering based approach. This proposed technique will be very helpful to deal with data sparsity problems; even when explicit trust rating data is unavailable. The finding will contribute in the area of recommender systems by improving the overall quality of automated recommendation making.