=Paper=
{{Paper
|id=Vol-1670/paper-51
|storemode=property
|title=None
|pdfUrl=https://ceur-ws.org/Vol-1670/paper-51.pdf
|volume=Vol-1670
}}
==None==
<pdf width="1500px">https://ceur-ws.org/Vol-1670/paper-51.pdf</pdf>
<pre>
       Discovering Data Transformations in Web
                 Resources (Abstract)

             Ziawasch Abedjan1  John Morcos2  Ihab F. Ilyas2
            Mourad Ouzzani Paolo Papotti Michael Stonebraker5
                          3             4

  1                2
      TU Berlin    University of Waterloo 3 Qatar Computing Research Institute
                    4                            5
                      Arizona State University     MIT CSAIL
                abedjan@tu-berlin.de {jmorcos,ilyas}@uwaterloo.ca
           mouzzani@qf.org.qa ppapotti@asu.edu stonebraker@csail.mit.edu



        Abstract. In data integration, data curation, and other data analy-
        sis tasks, users spend a considerable amount of time converting data
        from one representation to another. For example US dates to European
        dates or airport codes to city names. In practice, data scientists have
        to code most of the transformation tasks manually, search for the ap-
        propriate dictionaries, and involve domain experts. In a previous vision
        paper, we presented the initial design of DataXFormer, a system that
        uses web resources to assist in transformation discovery [1]. Specifically,
        DataXFormer discovers possible transformations from web tables and web
        forms and involves human feedback where appropriate. We demonstrated
        the system at SIGMOD 2015 and deployed an open version of the sys-
        tem, which helped us to increase our initial workload from 50 to 120
        transformations [3]. At the same time we extended DataXFormer with
        new algorithms to find
          1. transformations that entail multiple columns of input data,
          2. indirect transformations that are compositions of other transforma-
             tions,
          3. transformations that are not functions but rather relationships, and
          4. transformations from a knowledge base of public data.
        We report on experiments with the collection of 120 transformation tasks,
        and show our enhanced system automatically covers 101 of them by using
        openly available web resources [2].


References
1. Z. Abedjan, J. Morcos, M. Gubanov, I. F. Ilyas, M. Stonebraker, P. Papotti, and
   M. Ouzzani. DataXFormer: Leveraging the web for semantic transformations. In
   CIDR, 2015.
2. Z. Abedjan, J. Morcos, I. F. Ilyas, P. Papotti, M. Ouzzani, and M. Stonebraker.
   DataXFormer: A robust data transformation system. In ICDE, 2016.
3. J. Morcos, Z. Abedjan, I. F. Ilyas, M. Stonebraker, P. Papotti, and M. Ouzzani.
   DataXFormer: An interactive data transformation tool. In SIGMOD, 2015.

</pre>