Recommender systems based on collaborative filtering are vulnerable to shilling assaults owing to its openness. Shiller’s introduce fictitious profiles into the database of the system with the goal of modifying the recommendations to their advantage. To make product recommendations more reliable and robust, it is essential to find potential attackers concealed in recommender systems. Several shilling attack detection techniques have been proposed thus far, and they all have advantages and disadvantages that are complementary to different attacks. We present a thorough experimental evaluation of detection of attackers using K-nearest neighbour and Support Vector machine in this paper. Amazon electronic dataset is used to simulate several attack models. In the meantime, for investigation attackers are produced in our dataset using average-filling, random-filling and bandwagon attack models. For detection of shilling attack two models i.e., K Nearest neighbour and Support Vector Machine are also compared. Precision, recall, F1-score and accuracy are used to analysis the performance of these two algorithms. The result of our experiments shows that SVM is more accurate than KNN in detection of shilling attackers.