Basil Ell

Labels in the web of data

Entities on theWeb of Data need to have labels in order to be exposable to humans in a meaningful way. These labels can then be used for exploring the data, i.e., for displaying the entities in a linked data browser or other front-end applications, but also to support keywordbased or natural-language based search over the Web of Data. Far too many applications fall back to exposing the URIs of the entities to the user in the absence of more easily understandable representations such as human-readable labels. In this work we introduce a number of labelrelated metrics: completeness of the labeling, the efficient accessibility of the labels, unambiguity of labeling, and the multilinguality of the labeling. We report our findings from measuring the Web of Data using these metrics. We also investigate which properties are used for labeling purposes, since many vocabularies define further labeling properties beyond the standard property from RDFS.

Semantic mediawiki in operation: experiences with building a semantic portal

Wikis allow users to collaboratively create and maintain content. Semantic wikis, which provide the additional means to annotate the content semantically and thereby allow to structure it, experience an enormous increase in popularity, because structured data is more usable and thus more valuable than unstructured data. As an illustration of leveraging the advantages of semantic wikis for semantic portals, we report on the experience with building the AIFB portal based on Semantic MediaWiki. We discuss the design, in particular how free, wiki-style semantic annotations and guided input along a predefined schema can be combined to create a flexible, extensible, and structured knowledge representation. How this structured data evolved over time and its flexibility regarding changes are subsequently discussed and illustrated by statistics based on actual operational data of the portal. Further, the features exploiting the structured data and the benefits they provide are presented. Since all benefits have its costs, we conducted a performance study of the Semantic MediaWiki and compare it to MediaWiki, the nonsemantic base platform. Finally we show how existing caching techniques can be applied to increase the performance.

SPARTIQULATION: Verbalizing SPARQL Queries

Much research has been done to combine the fields of Databases and Natural Language Processing. While many works focus on the problem of deriving a structured query for a given natural language question, the problem of query verbalization – translating a structured query into natural language – is less explored. In this work we describe our approach to verbalizing SPARQL queries in order to create natural language expressions that are readable and understandable by the human day-to-day user. These expressions are helpful when having search engines that generate SPARQL queries for user-provided natural language questions or keywords. Displaying verbalizations of generated queries to a user enables the user to check whether the right question has been understood. While our approach enables verbalization of only a subset of SPARQL 1.1, this subset applies to 90 % of the \(209\) queries in our training set. These observations are based on a corpus of SPARQL queries consisting of datasets from the QALD-1 challenge and the ILD2012 challenge.

A language-independent method for the extraction of RDF verbalization templates

With the rise of the Semantic Web more and more data become available encoded using the Semantic Web standard RDF. RDF is faced towards machines: designed to be easily processable by machines it is difficult to be understood by casual users. Transforming RDF data into human-comprehensible text would facilitate non-experts to assess this information. In this paper we present a languageindependent method for extracting RDF verbalization templates from a parallel corpus of text and data. Our method is based on distant-supervised simultaneous multi-relation learning and frequent maximal subgraph pattern mining. We demonstrate the feasibility of our method on a parallel corpus of Wikipedia articles and DBpedia data for English and German.

Enterprise Knowledge Structures

One of the major aims of knowledge management has always been to facilitate the sharing and reuse of knowledge. Over the years a long list of technologies and tools pursuing this aim have been proposed, using different types of conceptual structures to capture the knowledge that individuals and groups communicate and exchange. This chapter is concerned with these knowledge structures and their development, maintenance and use within corporate environments. Enterprise knowledge management as we know it today often follows a predominantly community-driven approach to meet its organizational and technical challenges. It builds upon the power of mass collaboration and social software combined with intelligent machine-driven information management technology delivered though formal semantics. The knowledge structures underlying contemporary enterprise knowledge management platforms are diverse, from database tables deployed company-wide to files in proprietary formats used by scripts, from loosely defined folksonomies describing content through tags to highly formalized ontologies through which new enterprise knowledge can be automatically derived. Leveraging such structures requires a knowledge management environment which not only exposes them in an integrated fashion, but also allows knowledge workers to adjust and customize them according to their specific needs. We discuss how the Semantic MediaWiki provides such an environment – not only as an easy-to-use, highly versatile communication and collaboration medium, but also as an integration and knowledge engineering tool targeting the full range of enterprise knowledge structures currently used.

Learning a Cross-Lingual Semantic Representation of Relations Expressed in Text

Learning cross-lingual semantic representations of relations from textual data is useful for tasks like cross-lingual information retrieval and question answering. So far, research has been mainly focused on cross-lingual entity linking, which is confined to linking between phrases in a text document and their corresponding entities in a knowledge base but cannot link to relations. In this paper, we present an approach for inducing clusters of semantically related relations expressed in text, where relation clusters i can be extracted from text of different languages, ii are embedded in a semantic representation of the context, and iii can be linked across languages to properties in a knowledge base. This is achieved by combining multi-lingual semantic role labeling SRL with cross-lingual entity linking followed by spectral clustering of the annotated SRL graphs. With our initial implementation we learned a cross-lingual lexicon of relation expressions from English and Spanish Wikipedia articles. To demonstrate its usefulness we apply it to cross-lingual question answering over linked data.

SPARQL Query Verbalization for Explaining Semantic Search Engine Queries

In this paper we introduce Spartiqulation, a system that translates SPARQL queries into English text. Our aim is to allow casual end users of semantic applications with limited to no expertise in the SPARQL query language to interact with these applications in a more intuitive way. The verbalization approach exploits domain-independent template-based natural language generation techniques, as well as linguistic cues in labels and URIs.

Deriving human-readable labels from SPARQL queries

Over 80% of entities on the Semantic Web lack a human-readable label. This hampers the ability of any tool that uses linked data to offer a meaningful interface to human users. We argue that methods for deriving human-readable labels are essential in order to allow the usage of the Web of Data. In this paper we explore, implement, and evaluate a method for deriving human-readable labels based on the variable names used in a large corpus of SPARQL queries that we built from a set of log files. We analyze the structure of the SPARQL graph patterns and offer a classification scheme for graph patterns. Based on this classification, we identify graph patterns that allow us to derive useful labels. We also provide an overview over the current usage of SPARQL in the newly built corpus.

Semantically Enhanced Interactions between Heterogeneous Data Life-Cycles

This paper highlights how Semantic Web technologies facil- itate new socio-technical interactions between researchers and libraries focussing research data in a Virtual Research Environment. Concerning data practices in the elds of social sciences and humanities, the worlds of researchers and librarians have so far been separate. The increased digitization of research data and the ubiquitous use of Web technologies change this situation and oer new capacities for interaction. This is re- alized as a semantically enhanced Virtual Research Environment, which oers the possibility to align the previously disparate data life-cycles in research and in libraries covering a variety of inter-activities from import- ing research data via enriching research data and cleansing to exporting and sharing to allow for reuse. Currently, collaborative qualitative and quantitative analyses of a large digital corpus of educational lexica are carried out using this semantic and wiki-based research environment.