=Paper=
{{Paper
|id=Vol-1741/jist2016pd_paper3
|storemode=property
|title=An RDF Platform for Generating Web API for Open Government Data
|pdfUrl=https://ceur-ws.org/Vol-1741/jist2016pd_paper3.pdf
|volume=Vol-1741
|authors=Pattama Krataithong,Marut Buranarach,Nuttanont Hongwarittorrn,Thepchai Supnithi
|dblpUrl=https://dblp.org/rec/conf/jist/KrataithongBHS16
}}
==An RDF Platform for Generating Web API for Open Government Data==
https://ceur-ws.org/Vol-1741/jist2016pd_paper3.pdf
An RDF Platform for Generating Web API for Open
Government Data
Pattama Krataithong1,2, Marut Buranarach1, Nuttanont Hongwarittorrn2
and Thepchai Supnithi1
1
Language and Semantic Technology Laboratory
National Electronics and Computer Technology Center (NECTEC), Pathumthani, Thailand
{marut.bur,pattama.kra}@nectec.or.th
2
Department of Computer Science, Faculty of Science and Technology
Thammasat University, Pathumthani, Thailand
nth@cs.tu.ac.th
Abstract. Most of datasets in open data portals are mainly in tabular format in
spreadsheet, e.g. CSV and XLS. To increase the value and reusability of these
datasets, the datasets should be made available in RDF format that can support
better data querying and data integration. However, publishing and querying
RDF requires different knowledge and skills. In this poster, we present a
platform for publishing and querying the dataset in RDF that does not require
the user’s knowledge of RDF and SPARQL. This framework supports semi-
automatic construction of RDF data and RESTFul APIs from the datasets in
tabular format. The framework provides automatic schema detection, i.e. data
type detection, and ontology and RDF data mapping generation. RESTful API
is provided on top of the SPARQL data querying service for each published
RDF dataset. A platform prototype was developed and demonstrated using
some datasets from the Data.go.th website. Some current research directions
include automatic dataset API generation based on Web crawler and validator
and development of intelligent search engine over the dataset APIs.
Keywords: dataset management, open data platform, RDF data publishing
1 Introduction
The number of datasets on the Thailand open government data portal, i.e. Data.go.th,
is continually increasing. Majority of datasets on these portals are in tubular formats
such as Excel and CSV. Based on the 5-star open data model1, Resource Description
Framework (RDF) is a standard data format that can support linked open data. There
are two important standards for integrating data. First, RDF is a standard format for
integrating data based on URI and XML syntax. Second, the Web Ontology Language
(OWL) is important for linked data based on classes and properties.
1
http://5stardata.info/en/
Consuming RDF data is usually achieved by querying via an SPARQL endpoint. A
developer who wants to use the SPARQL endpoint must have the knowledge about
SPARQL and RDF. Our work proposes that Web API is an easier way for retrieving
RDF-based open data. There are several advantages of proving Web API over the
RDF datasets including:
Data as a service – developers who do not have background in RDF and
SPARQL can query a dataset via a RESTFul API service.
Standard data format– developers do not need to study a new data format, the
query results will be returned in the standard JSON format.
In this poster, we present a platform that provides a data management support for
RDF data publishing and consuming. The platform was developed using the Ontology
Application Management (OAM) framework [1]. The platform prototype was
available at the Demo-api.data.go.th website, which exemplifies deployment of the
platform using some datasets from the Data.go.th website.
2 RDF Dataset Management Process
Fig. 1a shows the RDF dataset publishing process. The RDF dataset generations
consist of four processes [2]: 1) User Management and Authentication 2) Dataset
Preparing and Import 3) Schema Detection and Verification and 4) OWL and RDF
data generation. The requirements of input data are as follows: 1) the dataset must
consist of only one table (one spreadsheet), 2) the table must have one header row, 3)
header of the table may be written in English or Thai language.
Fig. 1b shows the layers of Dataset Service API at Demo-api.data.go.th. The
website provides the data as a service through RESTFul APIs for each dataset which
was converted in RDF format and published on this portal. A form-based search
interface for each dataset is formed based on the ontology of the dataset. For each
dataset, the data querying service is automatically provided as RESTFul API by
means of the OAM framework. Application developers can query each dataset via
APIs and the returned search results are provided in JSON format.
a) RDF dataset publishing workflow b) RDF dataset service API
Fig. 1 RDF Dataset Management Process for Demo-api.data.go.th
3 Usage Scenarios
Fig. 2 shows an interface of the dataset publishing functions for each user. The user
can choose to create new, update or delete an RDF dataset. Fig. 3 shows an interface
of the schema detection and verification step. In accessing each created dataset the
user can choose to search or view the RDF dataset. Fig. 4 shows access to the dataset
API via the ontology-based search interface.
The APIs are provided for three main functions: getting all dataset names, getting
description of a dataset schema and querying a dataset [3]. An example of API for
querying a dataset by search conditions is shown in Fig. 5.
Fig. 2 User interface for listing all datasets and functions for each user
Fig. 3 User interface for the schema verification step
Fig. 4 Access to the API via the ontology-based search interface
Fig. 5 Example API for querying a dataset by search conditions
4 Discussion and Research Directions
Fig. 6 An automatic approach for generating APIs for the datasets of Data.go.th
This poster describes a semi-automatic framework for generating RDF dataset
from open tabular data. This framework allows the users to publish their datasets in
RDF format and query the data via Web API with no required knowledge about RDF
and SPARQL. One of the difficulties is that some datasets are not in the valid tabular
format [4]. In addition, human intervention is still required, which limited scalability
of the framework. One of our research directions is to develop a Web crawler and
validator to automatically retrieve and create the APIs from all valid datasets of the
Data.go.th website. We are also developing an intelligent search system, which allows
the user to search the data in the datasets via the APIs using a semi-natural-language
UI. The automatic approach for generating APIs for the datasets is shown in Fig 6.
References
1. Buranarach, M. et al.: OAM: An Ontology Application Management Framework for
Simplifying Ontology-Based Semantic Web Application Development. Int. J. Softw. Eng.
Knowl. Eng. 26, 01, 115–145 (2016).
2. Krataithong, P., Buranarach, M., Supnithi, T., and Hongwarittorrn, N.: Semi-Automatic
Framework for Generating RDF Dataset from Open Data. In: Proc. of the 11th
International Symposium on Natural Language Processing (SNLP2016) (2016).
3. Krataithong, P., Buranarach, M., and Supnithi, T.: RDF Dataset Management Framework
for Data.go.th. In: Proc. of the 10th International Conference on Knowledge, Information
and Creativity Support Systems (KICSS2015) (2015).
4. Ermilov I, Auer S, Stadler C (2013) User-driven semantic mapping of tabular data. Proc
9th Int Conf Semant Syst - I-SEMANTICS ’13 105. doi: 10.1145/2506182.2506196