Data Science Stack Exchange is a question and answer site for Data science professionals, Machine Learning specialists, and those interested in learning more about the field. It only takes a minute to sign up.
Sign up to join this community
I have been studying about information retrieval and recommender systems. While reading about it I found that accuracy not a useful measure in information retrieval.
I understand that, accuracy might not be good because if data imabalance, which is pretty common in information retrieval where we can have few data of our interest compared to the whole data.
But is there any other reason besides data imabalance for which we don't want to use accuracy for information retrieval problems?
Thanks!
Well, there are several reasons you can consider actually on why accuracy might not be the best measure of performance for an information retrieval system. For one, accuracy only considers the number of correct predictions made by the system, but not the relative importance of those predictions. For example, if a system has an accuracy of 90%, but only returns one relevant result out of a hundred, that result might not be very useful to the user.
Even considering that, accuracy is a global measure that does not provide any information about the performance of the system on specific queries or groups of queries. This can make it difficult to identify areas where the system is performing well or poorly, and to make targeted improvements to the system.
That's why Precision, recall, and F1 score are more useful measures for evaluating the performance.
You can also refer to https://nlp.stanford.edu/IR-book/html/htmledition/evaluation-of-unranked-retrieval-sets-1.html
