Evaluation of the results

Determining the value or worth of something in terms of quantity and quality is the process of evaluation. The increasing sophistication of text systems necessitates evaluation frameworks that measure the stated objectives and anticipated results. These frameworks serve a dual role – assessing different versions of the same product and also comparing similar systems. The topic of evaluation has grown into an essential part of systems development and a research field of its own.

Numerous convenient methods have been put forth to evaluate ML systems, which frequently make use of various computer- and human-centered metrics, most commonly known as objective and subjective evaluation. For example, using objective metrics allows us to measure something consistently and typically defies interpretation; either the spam detector achieved an accuracy above a threshold or didn’t. On the other hand, subjective evaluations are more expensive and time...