There is an increasing presence of structured data due to the adoption of Linked data principles on the web. At the same time, web users have di erent skills and want to be able to interact with Linked datasets in various manner, such as asking questions in natural language. This paper proposed a rst implementation of Query Answering system (QA) applied to the Linked Open Vocabularies (LOV) catalogue, mainly focused on metadata information retrieval. The goal is to provide to end users yet another access to metadata information available in LOV using natural language questions.
The recent years have seen the adoption of semantic web in many domains, generating a mass of structured data available in RDF. Linked data has contributed to interlink di erent datasets across domains, where publishers use best practices for producing interoperable datasets by reusing ontologies and making alignments.
However, users need to have a minimal knowledge of SPARQL language and
RDF skills to query RDF datasets. It is one barrier that is overcome by
Question Answering (QA) systems, which directly take as input questions in natural
language. The need for more advanced tools and QA systems that operate over
large repositories of Linked Data has also been the motivation for the QALD
Question Answering over Linked Data (QALD) series of workshops [
Vocabulary catalogues are special datasets where classes and properties are
used to model and generate datasets available in the Linked Data space. Most
vocabulary catalogues provide terms search and APIs to access their datasets.
LOV provides ve types of search methods: metadata search, ontology search,
APIs access, dump le in RDF and SPARQL endpoint access [
This paper presents a prototype for a vocabulary backed question answering system that can transform natural language questions into SPARQL queries, thus giving the end users access to the information stored in vocabulary repositories. The paper is structured as follows: Section 2 describes the system, followed by the set of questions in Section 3. An evaluation is presented in Section 4 and a short conclusion and future work in Section 5. 2
The system receives as input a question formulated in English and outputs the query that will retrieve the answer to the question from the LOV catalogue. The architecture of the system is illustrated in Fig. 1 and a screenshot of the live demo in Fig. 2. The system is available at http://lov.okfn.org/dataset/lov/qa.
The implementation uses the Quepy tool from Machinalis 3. The POS tagset
used by Quepy is the Penn Tagset [
QA4LOV: A Natural Language Interface to Linked Open Vocabulary
A vocabulary is de ned by a xed relation voaf:Vocabulary and a POS associated to a vann:preferredNamespacePrefix. LOV uses a unique pre x to identify namespaces, which is a string from length 2 to length 17, although the recommendation for publishers is to use a pre x with less than 10 characters. Our approach is based on attaching a regex to vocabulary represented by its pre x.
The syntactic processor is based on regular expressions using POS terms. As a vocabulary is identi ed by its pre x, we use the following syntactic patterns: NN, NNS, FW, DT, JJ and VBN. Each question from Q1 to Q14 is associated to a unique template. After, when a pre x is recognized, the semantic interpreter uses xed relations with the English tag, which are properties in RDF triple pattern. The code below shows a regex for the contributors of any vocabulary: LOV is a highly curated observatory of the semantic vocabularies ecosystem, with the aim of promoting the reuse of well-documented vocabularies in the Linked Data space. Each version of a vocabulary in LOV contains relevant metadata information which can be discovered by agents.
Additionally, users could be interested in other interesting facts, such as the number of versions, the number of datasets using the vocabulary, the number of external vocabularies reusing the vocabulary and the category to which belong the vocabulary. A rst set of 14 templates can be handled by the prototype, covering di erent type of metadata information available in a vocabulary. Table 1 shows the list of questions, where in the column \Template", [be] can be either the present or the past form of the verb, and [vocab] is the preferred pre x of the vocabulary. The regex uses lemmas to combine di erent types of the questions accordingly. 4
The system allows users to interact with the LOV catalogue through the answers generated. Depending on the types of the results (e.g., agents, versions, categories), the system allows users to further explore the dataset with more interactions. All the questions which generated SPARQL query gives satisfactory results. The most challenging issue is to determine the most suitable POS that cover all the vocabulary pre xes. For example, out of 5285 vocabularies in LOV, 13 of them contain an hyphen and the system can not generate query (e.g., 5 This number corresponds to the total number of vocabularies inserted in LOV as of
January, 8th 2016. elseweb-modelling). Moreover, 21 pre xes contain a number (e.g., g50k) and 3 special cases (homeActivity, LiMo, and juso.kr)and are not currently covered by the system. Currently, the system handles 92,99% of the pre xes in LOV. 5
In this paper, we have presented a prototype system for answering a set of questions in natural language backed by a vocabulary catalogue. Accessing LOV dataset by this system will hugely help lay users without SPARQL skills to interact more with the catalogue, and improve also the quality of the metadata by ontology publishers. The implementation uses the LOV dataset in RDF and the Quepy tool. The rst results show that the system handles 92,99% of the metadata of vocabularies in the LOV catalogue. We plan to extend the types of queries to more complex ones. Moreover, we can use various semantic relationships in LOV to do query expansion by using for instance sub-properties.