@inproceedings{proisl-uhrig-2016-somajo,
    title = "{S}o{M}a{J}o: State-of-the-art tokenization for {G}erman web and social media texts",
    author = "Proisl, Thomas  and
      Uhrig, Peter",
    editor = {Cook, Paul  and
      Evert, Stefan  and
      Sch{\"a}fer, Roland  and
      Stemle, Egon},
    booktitle = "Proceedings of the 10th Web as Corpus Workshop",
    month = aug,
    year = "2016",
    address = "Berlin",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/W16-2607/",
    doi = "10.18653/v1/W16-2607",
    pages = "57--62"
}