Towards NLP-supported Semantic Data Management

The heterogeneity of data poses a great challenge when data from different sources is to be merged for one application. Solutions for this are offered, for example, by ontology-based data management (OBDM). A challenge of OBDM is the automatic creation of semantic models from datasets. This process is typically performed either data- or label-driven and always involves manual human intervention. We identified textual descriptions of data, a form of metadata, quickly to be produced and consumed by humans, as third possible basis for automatic semantic modelling. In this paper, we present, how we plan to use textual descriptions to enhance semantic data management. We will use state of the art NLP technologies to identify concepts within textual descriptions and build semantic models from this in combination with an evolving ontology. We will use automatically identified models in combination with the human data provider to automatically extend the ontology so that it learns new verified concepts over time. Finally, we will use the created ontology and automatically identified semantic models to either rate descriptions for new data sources or even to automatically generate descriptive texts that are easier to understand by the human user than formal models. We present the procedure which we plan for the ongoing research, as well as expected outcomes.

[1]  Jingqi Wang,et al.  Enhancing Clinical Concept Extraction with Contextual Embedding , 2019, J. Am. Medical Informatics Assoc..

[2]  Dieter Fensel,et al.  Knowledge Engineering: Principles and Methods , 1998, Data Knowl. Eng..

[3]  Ian Horrocks,et al.  Optique 1.0: Semantic Access to Big Data: The Case of Norwegian Petroleum Directorate's FactPages , 2013, International Semantic Web Conference.

[4]  Ravi Lourdusamy,et al.  A Survey on Methods of Ontology Learning from Text , 2019 .

[5]  Thomas Schauppenlehner,et al.  Theoretical Availability versus Practical Accessibility: The Critical Role of Metadata Management in Open Data Portals , 2018 .

[6]  Kristina Lerman,et al.  Semi-automatically Mapping Structured Sources into the Semantic Web , 2012, ESWC.

[7]  Maurizio Lenzerini.  Ontology-based data management , 2011, CIKM '11.

[8]  Mohammed Bennamoun,et al.  Ontology learning from text: A look back and into the future , 2012, CSUR.

[9]  Sören Auer,et al.  Towards Cleaning-Up Open Data Portals: A Metadata Reconciliation Approach , 2015, 2016 IEEE Tenth International Conference on Semantic Computing (ICSC).

[10]  Craig A. Knoblock,et al.  Learning Semantic Models of Data Sources Using Probabilistic Graphical Models , 2019, WWW.

[11]  Tobias Meisen,et al.  ESKAPE: Information Platform for Enabling Semantic Data Processing , 2017, ICEIS.

[12]  Diego Klabjan,et al.  Automatic Ontology Learning from Domain-Specific Short Unstructured Text Data , 2019, KMIS.

[13]  Tobias Meisen,et al.  You are Missing a Concept! Enhancing Ontology-Based Data Access with Evolving Ontologies , 2019, 2019 IEEE 13th International Conference on Semantic Computing (ICSC).

[14]  Ming-Wei Chang,et al.  BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding , 2019, NAACL.

[15]  Yiming Yang,et al.  XLNet: Generalized Autoregressive Pretraining for Language Understanding , 2019, NeurIPS.

[16]  Marijn Janssen,et al.  The potential of metadata for linked open data and its value for users and publishers , 2012 .

[17]  Zhaohui Wu,et al.  Using Prerequisites to Extract Concept Maps fromTextbooks , 2016, CIKM.

[18]  Luke S. Zettlemoyer,et al.  Deep Contextualized Word Representations , 2018, NAACL.

[19]  Steffen Staab,et al.  Ontology Learning for the Semantic Web , 2002, IEEE Intell. Syst..