🔡 N-gram Language Modeling

< Back to Projects

< Back Home

N-gram Student Model

Figure 1: The results of a trained N-gram model on my teammates' and my (students') corpus of code.

N-gram Teacher Model

Figure 2: The results of a trained N-gram model on my instructor's corpus of code.

This project explores code completion in Java, leveraging N-gram language modeling. The N-gram model predicts the next token in a sequence by learning the probability distributions of token occurrences in training data. The model selects the most probable token based on learned patterns, making it a fundamental technique in natural language processing and software engineering automation.

Source Code