Link Prediction is very important aspect of Social Network Analysis. Many popular applications like Facebook and LinkedIn are using Link Prediction as its feature. In the co-authorship network, we are interested in Link Prediction for Scholar Social Networks. Associations among the researchers from the network of links which can be termed as co-authorship network. For fruitful results in research area, association between authors is most important than individual work. If we have such data set which consists more information about the author and about the research paper then there will be more accuracy in predicting the links. In this work, we proposed a technique which is based on node similarity measure and Artificial Neural Network concept. We consider the features or information which are available in data set then assign the score to all pairs of nodes and apply Artificial Neural Network concept to the system. The proposed technique will improve the accuracy of Link Prediction.