TaPP 2020 Data Provenance for Attributes: Attribute Lineage Dennis Dosso, Susan Davidson, Gianmaria Silvello dosso@dei.unipd.it (University of Padua, Italy) susan@seas.upenn.edu (University of Pennsylvania, USA) silvello@dei.unipd.it (University of Padua, Italy)
Overview • Data Provenance & Lineage • An Example • Conclusions and Future Work TaPP 2020 Dennis Dosso, Susan Davidson, Gianmaria Silvello 2
Data Provenance & Lineage TaPP 2020 Dennis Dosso, Susan Davidson, Gianmaria Silvello 3
Data Provenance & Lineage Data Provenance is a form of metadata describing the life of data • A relational database instance I • A query Q • An output Q(I) • A tuple t in Q(I) The lineage of t is the set of all and only the tuples in the input that were used for the generation of the output We introduce a new type a lineage defined for attributes The idea: it is the set of all the attributes in the input that are used by the query to produce an attribute in a tuple of the output What we want to do: define lineage for a attribute value in the output It has a di ff erent level of granularity, and can help in tasks such as finding attribute hotspots in I TaPP 2020 Dennis Dosso, Susan Davidson, Gianmaria Silvello 4
An Example TaPP 2020 Dennis Dosso, Susan Davidson, Gianmaria Silvello 5
<latexit sha1_base64="pEOx6SD2QtN9xs1BJXpZn0aNTY=">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</latexit> <latexit sha1_base64="GNR/jYPDCtA25phgY9V+aQUXxB0=">AB+HicbVC7TsNAEDyHVwivACXNiQSJKrITiVcVQUOZSOQhJVF0vmzCkfPZulsjBSv/QAsVHaLlbyj4F2xjIV5TjWZ2tbPjBlIYtO03K7ewuLS8kl8trK1vbG4Vt3faxg81hxb3pa+7LjMghYIWCpTQDTQwz5XQcacXid+5BW2Er65wFsDAYxMlxoIzjKV2uTl0zsrDYsmu2CnoX+JkpEQyNIbF9/7I56EHCrlkxvQcO8BxDQKLmFe6IcGAsanbAK9mCrmgRlEado5PQgNQ58GoKmQNBXh+0bEPGNmnhtPegyvzW8vEf/zeiGOTwaRUEGIoHhyCIWE9JDhWsQ1AB0JDYgsSQ5UKMqZoigBWcx2IY91JI+zhNcPT1/V/SrlacWqXWrJbq51kzebJH9skhcgxqZNL0iAtwskNuScP5NG6s56sZ+vlczRnZTu75Aes1w/JhJLT</latexit> <latexit sha1_base64="jvFCPOYHGD7LNqmV1Keq3WLRhBs=">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</latexit> <latexit sha1_base64="/LpCJpmSOSZpapK49YkMXhSiSlc=">ACEHicbVC7SgNBFJ2NrxhfUstBoMQcJuImgjBG2sJAbzgCQs5ObZMjs7DJzVwhLGj/Br7DVyk5s/QML/8VNTKHRUx3OuZdz7/FCKQza9oeVWlhcWl5Jr2bW1jc2t7LbO3UTRJpDjQcy0E2PGZBCQ0FSmiGpjvSWh4w8uJ37gDbUSgbnEUQsdnfSV6gjNMJDe7X3VL9Jy2Q+HG18yHY1oZBArG+apbPHKzObtgT0H/EmdGcmSGipv9bHcDHvmgkEtmTMuxQ+zETKPgEsaZdmQgZHzI+tBKqEoCTSefjGmh5FhGNAQNBWSTkX4uREz35iR7yWTPsOBmfcm4n9eK8LeWScWKowQFJ8EoZAwDTJci6QeoF2hAZFNLgcqFOVM0TQgjLOEzFK+sokfTjz3/8l9WLBKRWKNye58sWsmTZIwckTxySsrkilRIjXByTx7JE3m2HqwX69V6+x5NWbOdXfIL1vsXiSbNA=</latexit> <latexit sha1_base64="pEOx6SD2QtN9xs1BJXpZn0aNTY=">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</latexit> <latexit sha1_base64="GNR/jYPDCtA25phgY9V+aQUXxB0=">AB+HicbVC7TsNAEDyHVwivACXNiQSJKrITiVcVQUOZSOQhJVF0vmzCkfPZulsjBSv/QAsVHaLlbyj4F2xjIV5TjWZ2tbPjBlIYtO03K7ewuLS8kl8trK1vbG4Vt3faxg81hxb3pa+7LjMghYIWCpTQDTQwz5XQcacXid+5BW2Er65wFsDAYxMlxoIzjKV2uTl0zsrDYsmu2CnoX+JkpEQyNIbF9/7I56EHCrlkxvQcO8BxDQKLmFe6IcGAsanbAK9mCrmgRlEado5PQgNQ58GoKmQNBXh+0bEPGNmnhtPegyvzW8vEf/zeiGOTwaRUEGIoHhyCIWE9JDhWsQ1AB0JDYgsSQ5UKMqZoigBWcx2IY91JI+zhNcPT1/V/SrlacWqXWrJbq51kzebJH9skhcgxqZNL0iAtwskNuScP5NG6s56sZ+vlczRnZTu75Aes1w/JhJLT</latexit> <latexit sha1_base64="NMu3ASIFDtGrITIpScTj9rm7uEc=">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</latexit> <latexit sha1_base64="s9u24BDVkbr6qz4UjOZDK9EJoi4=">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</latexit> <latexit sha1_base64="jvFCPOYHGD7LNqmV1Keq3WLRhBs=">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</latexit> <latexit sha1_base64="/LpCJpmSOSZpapK49YkMXhSiSlc=">ACEHicbVC7SgNBFJ2NrxhfUstBoMQcJuImgjBG2sJAbzgCQs5ObZMjs7DJzVwhLGj/Br7DVyk5s/QML/8VNTKHRUx3OuZdz7/FCKQza9oeVWlhcWl5Jr2bW1jc2t7LbO3UTRJpDjQcy0E2PGZBCQ0FSmiGpjvSWh4w8uJ37gDbUSgbnEUQsdnfSV6gjNMJDe7X3VL9Jy2Q+HG18yHY1oZBArG+apbPHKzObtgT0H/EmdGcmSGipv9bHcDHvmgkEtmTMuxQ+zETKPgEsaZdmQgZHzI+tBKqEoCTSefjGmh5FhGNAQNBWSTkX4uREz35iR7yWTPsOBmfcm4n9eK8LeWScWKowQFJ8EoZAwDTJci6QeoF2hAZFNLgcqFOVM0TQgjLOEzFK+sokfTjz3/8l9WLBKRWKNye58sWsmTZIwckTxySsrkilRIjXByTx7JE3m2HqwX69V6+x5NWbOdXfIL1vsXiSbNA=</latexit> <latexit sha1_base64="s9u24BDVkbr6qz4UjOZDK9EJoi4=">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</latexit> Example Agencies tId Name BasedIn phone BayTours ( a 1 , 1 ) San Francisco ( a 1 , 2 ) 415-1200 ( a 1 , 3 ) a 1 HarborCruz ( a 2 , 1 ) Santa Cruz ( a 2 , 2 ) 831-3000 ( a 2 , 3 ) a 2 tId Name Destination type price ExternalTours BayTours ( e 1 , 1 ) San Francisco ( e 1 , 2 ) car ( e 1 , 3 ) 50 ( e 1 , 4 ) e 1 BayTours ( e 2 , 1 ) Santa Cruz ( e 2 , 2 ) bus ( e 2 , 3 ) 100 ( e 2 , 4 ) e 2 BayTours ( e 3 , 1 ) Santa Cruz ( e 3 , 2 ) boat ( e 3 , 3 ) 250 ( e 3 , 4 ) e 3 BayTours ( e 4 , 1 ) Monterey ( e 4 , 2 ) boat ( e 4 , 3 ) 400 ( e 4 , 4 ) e 4 HarborCruz ( e 5 , 1 ) Monterey ( e 5 , 2 ) boat ( e 5 , 3 ) 200 ( e 5 , 4 ) e 5 HarborCruz ( e 6 , 1 ) Carmel ( e 6 , 2 ) train ( e 6 , 3 ) 90 ( e 6 , 4 ) e 6 Q 1 : SELECT DISTINCT a.name, a.phone FROM Agencies a, ExternalTours e WHERE a.name = e.name AND e.type = ’boat’ tId Name Phone BayTours ( o 1 , 1 ) 415-1200 ( o 1 , 2 ) o 1 HarborCruz ( o 2 , 1 ) 831-3000 ( o 2 , 2 ) o 2 Example taken from: Cheney J, Chiticariu L, Tan WC. Provenance in databases: Why, how, and where. Now Publishers Inc; 2009. TaPP 2020 Dennis Dosso, Susan Davidson, Gianmaria Silvello 6
Recommend
More recommend
