FireCite: Lightweight real-time reference string extraction from webpages

We present FireCite, a Mozilla Firefox browser extension that helps scholars assess and manage scholarly references on the web by automatically detecting and parsing such reference strings in real-time. FireCite has two main components: 1) a reference string recognizer that has a high recall of 96%, and 2) a reference string parser that can process HTML web pages with an overall F1 of .878 and plaintext reference strings with an overall F1 of .97. In our preliminary evaluation, we presented our FireCite prototype to four academics in separate unstructured interviews. Their positive feedback gives evidence to the desirability of FireCite’s citation management capabilities.