Generating taxonomic terms for software bug classification by utilizing topic models based on Latent Dirichlet Allocation

Discovering categorical (taxonomic) terms in text classification is an important and complex problem. Development of a good text classifier depends on the method of identification and generation of proper taxonomic terms. Software bug indicates improper behavior of the functionalities given during the requirements. These bugs are tracked with the help of bug tracking systems (BTS) where the bug information is presented using several attributes out of which some important attributes are textual for example summary and description. For effective classification of the software bugs a good text classifying mechanism is required for which proper taxonomic terms are required to be identified. In this work a methodology is presented to find the taxonomic terms using Latent Dirichlet Allocation (LDA) for software bug classification.