Stakeholders of a software system usually deliver bug reports to a software bug tracking system to report problems they encounter during the use of that system. After that, those incoming requests are assigned to the proper technicians to be analysed and fixed. However, issue reporters frequently misclassify bug requests as non-bug and vice versa. This problem is called bug report misclassification. In fact, it is very costly in terms of time and effort as developers have to manually reclassify those requests to then be able to take the appropriate action. Accordingly, the automatic classification of incoming new reports would be of a valuable demanded feature in bug tracking systems. Careful analysis for literature related to this problem was carried out. It has been found that a hybrid approach combining between topic modelling as a feature extraction technique and the fuzzy logic as a classification technique is promising but rarely utilized approach. In this work, a combined approach of topic modelling and fuzzy logic was introduced to classify bug reports to bugs and non-bugs. The proposed approach was validated using three open-source projects. Finally, the conducted experiments have shown that proposed classifier achieves significant and competitive predictive accuracy.