Search

The article discusses importance of information retrieval system to meet information needs of users and mentions challenges of predicting normalized discounted cumulative gain (nDCG) methods. Topics discussed include data collection and judgements of test collections and dataset creation, creation of research by computing performance measures and use of linear regression model from creation of voting ensemble.

The article explores correlation and prediction of high-cost information retrieval evaluation metrics using deep learning. It focuses on reducing cost of the evaluation of information retrieval systems and the study proposes a method that employs deep learning to predict the precision evaluation metric. It presents method and the justifications for the correlations between some pairs of retrieval metrics.