Natural Language Explanations for Query Results

Multiple lines of research have developed Natural Language (NL) interfaces for formulating database queries. We build upon this work, but focus on presenting a highly detailed form of the answers in NL. The answers that we present are importantly based on the provenance of tuples in the query result, detailing not only the results but also their explanations. We develop a novel method for transforming provenance information to NL, by leveraging the original NL query structure. Furthermore, since provenance information is typically large and complex, we present two solutions for its effective presentation as NL text: one that is based on provenance factorization, with novel desiderata relevant to the NL case, and one that is based on summarization.

[1]  Boris Glavic.  Big Data Provenance: Challenges and Implications for Benchmarking , 2012, WBDB.

[2]  Khaled M. Elbassioni,et al.  On the readability of monotone Boolean formulae , 2011, J. Comb. Optim..

[3]  Chris Brew,et al.  TR Discover: A Natural Language Interface for Querying and Analyzing Interlinked Datasets , 2015, International Semantic Web Conference.

[4]  Juliana Freire,et al.  Provenance and scientific workflows: challenges and opportunities , 2008, SIGMOD Conference.

[5]  Val Tannen,et al.  Querying data provenance , 2010, SIGMOD Conference.

[6]  Susan B. Davidson,et al.  Addressing the provenance challenge using ZOOM , 2008, Concurr. Comput. Pract. Exp..

[7]  Daniel Deutch,et al.  Selective Provenance for Datalog Programs Using Top-K Queries , 2015, Proc. VLDB Endow..

[8]  Michael Clausen,et al.  Algebraic complexity theory , 1997, Grundlehren der mathematischen Wissenschaften.

[9]  Val Tannen,et al.  Provenance semirings , 2007, PODS.

[10]  Edith Hemaspaandra,et al.  Minimization for Generalized Boolean Formulas , 2011, IJCAI.

[11]  Adriane Chapman,et al.  Efficient provenance storage , 2008, SIGMOD Conference.

[12]  Yogesh L. Simmhan,et al.  Karma2: Provenance Management for Data-Driven Workflows , 2008, Int. J. Web Serv. Res..

[13]  Tova Milo,et al.  A Natural Language Interface for Querying General and Individual Knowledge , 2015, Proc. VLDB Endow..

[14]  Claire Gardent,et al.  Quelo Natural Language Interface: Generating queries and answer descriptions , 2014 .

[15]  Dan Suciu,et al.  A formal approach to finding explanations for database queries , 2014, SIGMOD Conference.

[16]  Daniel Deutch,et al.  Approximated Summarization of Data Provenance , 2015, CIKM.

[17]  Jakub Závodný,et al.  Factorised representations of query results: size bounds and readability , 2012, ICDT '12.

[18]  M. Emms.  Variants of Tree Similarity in a Question Answering Task , 2006 .

[19]  Andreas Haeberlen,et al.  Querying Provenance for Ranking and Recommending , 2012, TaPP.

[20]  Fei Li,et al.  Constructing an Interactive Natural Language Interface for Relational Databases , 2014, Proc. VLDB Endow..

[21]  Melanie Herschel,et al.  Provenance: On and Behind the Screens , 2016, SIGMOD Conference.