Renata Queiroz Dividino

Querying for provenance, trust, uncertainty and other meta knowledge in RDF

The Semantic Web is based on accessing and reusing RDF data from many different-sources, which one may assign different levels of authority and credibility. Existing Semantic Web query languages, like SPARQL, have targeted the retrieval, combination and re-use of facts, but have so far ignored all aspects of meta knowledge, such as origins, authorship, recency or certainty of data. In this paper, we present an original, generic, formalized and implemented approach for managing many dimensions of meta knowledge, like source, authorship, certainty and others. The approach re-uses existing RDF modeling possibilities in order to represent meta knowledge. Then, it extends SPARQL query processing in such a way that given a SPARQL query for data, one may request meta knowledge without modifying the query proper. Thus, our approach achieves highly flexible and automatically coordinated querying for data and meta knowledge, while completely separating the two areas of concern.

An Investigation of HTTP Header Information for Detecting Changes of Linked Open Data Sources

Data on the Linked Open Data (LOD) cloud changes frequently. Applications that operate on local caches of Linked Data need to be aware of these changes. In this way they can update their cache to ensure operating on the most recent version of the data. Given the HTTP basis recommended in the Linked Data guidelines, the native way of detecting changes would be to use HTTP header information, such as the Last-Modified field. However, it is uncertain to which degree this field is currently supported on the LOD cloud and how reliable the provided information is. In this paper, we analyse a large-scale dataset obtained from the LOD cloud by weekly crawls over almost two years. On these weekly snapshots, we observed that for only 15 % of the Linked Data resources the HTTP header field Last-Modified is actually available and that the date provided for the last modification aligns in only 8 % with the observed changes of the data itself.

Ranking RDF with provenance via preference aggregation

Information retrieval on RDF data benefits greatly from additional provenance information attached to the individual pieces of information. Provenance information such as origin of data, certainty, and temporal information on RDF statements can be used to rank search results according to one of those dimensions. In this paper, we consider the problem of aggregating provenance information from different dimensions in order to obtain a joint ranking over all dimensions. We relate this to the problem of preference aggregation in social choice theory and translate different solutions for preference aggregation to the problem of aggregating provenance rankings. By exploiting the ranking orderings on the provenance dimensions, we characterize three different approaches for aggregating preferences, namely the lexicographical rule, the Borda rule and the plurality rule, in our framework of provenance aggregation.

Strategies for Efficiently Keeping Local Linked Open Data Caches Up-To-Date

Quite often, Linked Open Data (LOD) applications pre-fetch data from the Web and store local copies of it in a cache for faster access at runtime. Yet, recent investigations have shown that data published and interlinked on the LOD cloud is subject to frequent changes. As the data in the cloud changes, local copies of the data need to be updated. However, due to limitations of the available computational resources (e.g., network bandwidth for fetching data, computation time) LOD applications may not be able to permanently visit all of the LOD sources at brief intervals in order to check for changes. These limitations imply the need to prioritize which data sources should be considered first for retrieving their data and synchronizing the local copy with the original data. In order to make best use of the resources available, it is vital to choose a good scheduling strategy to know when to fetch data of which data source. In this paper, we investigate different strategies proposed in the literature and evaluate them on a large-scale LOD dataset that is obtained from the LOD cloud by weekly crawls over the course of three years. We investigate two different setups: (i) in the single step setup, we evaluate the quality of update strategies for a single and isolated update of a local data cache, while (ii) the iterative progression setup involves measuring the quality of the local data cache when considering iterative updates over a longer period of time. Our evaluation indicates the effectiveness of each strategy for updating local copies of LOD sources, i. e, we demonstrate for given limitations of bandwidth, the strategies’ performance in terms of data accuracy and freshness. The evaluation shows that the measures capturing change behavior of LOD sources over time are most suitable for conducting updates.

Integrating business process and user interface models using a model-driven approach

Business services are complex entities that encompass descriptions about different aspects including business processes and user interfaces. Typically, the modeling of business services results on several correlated models. On the one hand, there is the need to keep these models apart in order to attain the levels of abstractions needed to model different aspects of a service. On the other hand, it is important to maintain the consistency and integrity of models. In this paper, we show how to maintain the consistency and integrity of models and present a use case for the integration of user interface design and business process models.