Our Project
We built an autocomplete system that is able to handle large amounts of data while still remaining fast, resouce-efficient, and accurate. Our project uses ngram data from the Google Books corpus, moving it through a pipeline that included pre-processing the data, storing and indexing it, querying it, and making it available via a web API. We developed a ranking system that determines the best completions to return based on the given text and the data available.