The folders description:

	1. Consists of experiments developed with the complete dataset of papers, using the full text of each paper. The key phrase extraction methods experimented here are: (1) Centrality Measures - Degree, Strength, Closeness, PageRank, Betweenness. (2) Statistics Measures - TF, TF-IDF.

	2. Consists of experiments developed with the complete dataset of papers, using exclusively the abstract of each paper. The key phrase extraction methods experimented here are: (1) Centrality Measures - Degree, Strength, Closeness, PageRank, Betweenness. (2) Statistics Measures - TF, TF-IDF.

The dataset description:

	1. Download from: http://disi.unitn.it/~krapivin/
	2. Consists of pairs of files, <id>.txt with full text, <id>.key with keyphrases each on a new line.
	3. There is the file "!authors.dat" inside the archive, it contains all authors for all papers.
	4. How to cite? Use the bibtex:

    @TECHREPORT{key:dataset2009krapivin-autayeu-marchese,
     AUTHOR =       {Mikalai Krapivin and Aliaksandr Autayeu and Maurizio Marchese},
     TITLE =        {Large Dataset for Keyphrases Extraction},
     INSTITUTION =  {DISI, Trento, Italy},
     YEAR =         {2008},
     month =        {May},
     number =       {DISI-09-055},
     note =         {http://eprints.biblio.unitn.it/archive/00001671/01/disi09055-krapivin-autayeu-marchese.pdf},
     url =         {\url{http://eprints.biblio.unitn.it/archive/00001671/01/disi09055-krapivin-autayeu-marchese.pdf}}
    }