PubTag: Generating Research Tag-Clouds with Keyphrase Extraction and Learning-to-Rank
- A labelled dataset of 1,126 author-scored keyphrases extracted from the text of papers coauthored by 12 Computer Science professors, which can be found in data/scored_keyphrases.
- An evaluation of four learning-to-rank frameworks using the above dataset, and a comparison of their results with an unsupervised keyphrase extraction framework, which can be found in data/evaluation
- Raw text data used for evaluation in data/txt