Segmentation of User Task Behavior by using Artificial Neural Network

As Segmentation of User's Task to understand the user search behavior is the new field of research for various researchers. Massive volumes of search log data have been collected in several search engines. Currently, a commercial search engine collects billions of queries and gathers terabytes of log data on each single day. At times user moves from one site to another because latency time of the site is more, so the researchers found this as an essential subject for research. Proposed work classifies the user query by combining query clustering boundary spread method with the neural network. For training of neural network proposed work evolve binary feature vector from the clustered query obtained from QCBSP method. The experiment was done on user search behavior of different time intervals. Results show that proposed work has achieved a high precision, accuracy for classification of the user query. Proposed scheme reduces execution time as well because of using trained neural network.