Semantic Web: – Interoperability,Usability,Applicability

Knowledge graphs (KGs) are a key ingredient to complement search results, discover entities and their relations and support several knowledge discovery tasks. We face the problem of building relatedness explanations, that is, graphs that can explain how a pair of entities is related in a KG. Explanations can be used in a variety of tasks; from exploratory search to query answering. We formalize the notion of explanation and present two algorithms. The first, E4D (Explanations from Data), assembles explanations starting from all paths interlinking the source and target entity in the data. The second algorithm E4S (Explanations from Schema) builds explanations focused on a specific relatedness perspective expressed by providing a predicate. E4S first generates candidate explanation patterns at the level of schema; then, it assembles explanations by proceeding to their verification in the data. Given a set of paths, found by E4D or E4S, we describe different criteria to build explanations based on information-theory, diversity, and their combination. As a concrete use-case of relatedness explanations, we introduce relatedness-based KG querying, which revisits the query-by-example paradigm from the perspective of relatedness explanations. We implemented all the machinery in the RECAP tool, which is based on RDF and SPARQL. We discuss an evaluation of the explanation building algorithms and a comparison of RECAP with related systems on real-world data.

Word embeddings are powerful for many tasks in natural language processing. In this work, we learn word embeddings using weighted graphs from word association norms (WAN) with the node2vec algorithm. Although building WAN is a difficult and time-consuming task, training the vectors from these resources is a fast and efficient process. This allows us to obtain good quality word embeddings from small corpora. We evaluate our word vectors in two ways: intrinsic and extrinsic. The intrinsic evaluation was performed with several word similarity benchmarks, WordSim-353, MC30, MTurk-287, MEN-TR-3k, SimLex-999, MTurk-771 and RG-65, and different similarity measures achieving better results than those obtained with word2vec, GloVe, and fastText, trained on a huge corpus. The extrinsic evaluation was done by measuring the quality of sentence embeddings using transfer tasks: sentiment analysis, paraphrase detection, natural language inference, and semantic textual similarity. The word vectors learned from the WAN are available on our Github page.

Current retrieval and recommendation approaches rely on hard-wired data models. This hinders personalized customizations to meet information needs of users in a more flexible manner. Therefore, the paper investigates how similarity-based retrieval strategies can be combined with graph queries to enable users or system providers to explore repositories in the Linked Open Data (LOD) cloud more thoroughly. For this purpose, we developed novel content-based recommendation approaches. They rely on concept annotations of Simple Knowledge Organization System (SKOS) vocabularies and a SPARQL-based query language that facilitates advanced and personalized requests for openly available knowledge graphs. We have comprehensively evaluated the novel search strategies in several test cases and example application domains (i.e., travel search and multimedia retrieval). The results of the web-based online experiments showed that our approaches increase the recall and diversity of recommendations or at least provide a competitive alternative strategy of resource access when conventional methods do not provide helpful suggestions. The findings may be of use for Linked Data-enabled recommender systems (LDRS) as well as for semantic search engines that can consume LOD resources.

One of the key requirements to facilitate the semantic analytics of information regarding contemporary and historical events on the Web, in the news and in social media is the availability of reference knowledge repositories containing comprehensive representations of events, entities and temporal relations. Existing knowledge graphs, with popular examples including DBpedia, YAGO and Wikidata, focus mostly on entity-centric information and are insufficient in terms of their coverage and completeness with respect to events and temporal relations. In this article we address this limitation, formalise the concept of a temporal knowledge graph and present its instantiation – EventKG. EventKG is a multilingual event-centric temporal knowledge graph that incorporates over 690 thousand events and over 2.3 million temporal relations obtained from several large-scale knowledge graphs and semi-structured sources and makes them available through a canonical RDF representation. Whereas popular entities often possess hundreds of relations within a temporal knowledge graph such as EventKG, generating a concise overview of the most important temporal relations for a given entity is a challenging task. In this article we demonstrate an application of EventKG to biographical timeline generation, where we adopt a distant supervision method to identify relations most relevant for an entity biography. Our evaluation results provide insights on the characteristics of EventKG and demonstrate the effectiveness of the proposed biographical timeline generation method.

Knowledge graphs, which contain annotated descriptions of entities and their interrelations, are often generated using rules that apply semantic annotations to certain data sources. (Re)using ontology terms without adhering to the axioms defined by their ontologies results in inconsistencies in these graphs, affecting their quality. Methods and tools were proposed to detect and resolve inconsistencies, the root causes of which include rules and ontologies. However, these either require access to the complete knowledge graph, which is not always available in a time-constrained situation, or assume that only generation rules can be refined but not ontologies. In the past, we proposed a rule-driven method for detecting and resolving inconsistencies without complete knowledge graph access, but it requires a predefined set of refinements to the rules and does not guide users with respect to the order the rules should be inspected. We extend our previous work with a rule-driven method, called Resglass, that considers refinements for generation rules as well as ontologies. In this article, we describe Resglass, which includes a ranking to determine the order with which rules and ontology elements should be inspected, and its implementation. The ranking is evaluated by comparing the manual ranking of experts to our automatic ranking. The evaluation shows that our automatic ranking achieves an overlap of 80% with experts ranking, reducing this way the effort required during the resolution of inconsistencies in both rules and ontologies.

Knowledge Graphs (KGs) have become useful sources of structured data for information retrieval and data analytics tasks. Enabling complex analytics, however, requires entities in KGs to be represented in a way that is suitable for Machine Learning tasks. Several approaches have been recently proposed for obtaining vector representations of KGs based on identifying and extracting relevant graph substructures using both uniform and biased random walks. However, such approaches lead to representations comprising mostly popular, instead of relevant, entities in the KG. In KGs, in which different types of entities often exist (such as in Linked Open Data), a given target entity may have its own distinct set of most relevant nodes and edges. We propose specificity as an accurate measure of identifying most relevant, entity-specific, nodes and edges. We develop a scalable method based on bidirectional random walks to compute specificity. Our experimental evaluation results show that specificity-based biased random walks extract more meaningful (in terms of size and relevance) substructures compared to the state-of-the-art and the graph embedding learned from the extracted substructures perform well against existing methods in common data mining tasks.

In this system paper we describe metaphactory, a platform for building knowledge graph applications. The metaphactory platform aims at supporting different categories of knowledge graph users within the organization by realizing relevant services for knowledge graph data management tasks, providing a rich and customizable user interface, and enabling rapid building of use case-specific applications. The paper discusses how the platform architecture design built on open standards enables its reusability in various application domains and use cases as well as facilitates integration of the knowledge graph with other parts of the organizational data and software infrastructure. We highlight the capabilities of the platform by describing its usage in four different knowledge graph application domains and share the lessons learnt from the practical experience of building knowledge graph applications in the enterprise context.

The increasing number of Knowledge Graphs (KGs) available today calls for powerful query languages that can strike a balance between expressiveness and complexity of query evaluation, and that can be easily integrated into existing query processing infrastructures. We present Extended Property Paths (EPPs), a significant enhancement of Property Paths (PPs), the navigational core included in the SPARQL query language. We introduce the EPPs syntax, which allows to capture in a succinct way a larger class of navigational queries than PPs and other navigational extensions of SPARQL, and provide formal semantics. We describe a translation from non-recursive EPPs (nEPPs) into SPARQL queries and provide novel expressiveness results about the capability of SPARQL sub-languages to express navigational queries. We prove that the language of EPPs is more expressive than that of PPs; using EPPs within SPARQL allows to express things that cannot be expressed when only using PPs. We also study the expressiveness of SPARQL with EPPs in terms of reasoning capabilities. We show that SPARQL with EPPs is expressive enough to capture the main RDFS reasoning functionalities and describe how a query can be rewritten into another query enhanced with reasoning capabilities. We complement our contributions with an implementation of EPPs as the SPARQL-independent iEPPs language and an implementation of the translation of nEPPs into SPARQL queries. What sets our approach apart from previous research on querying KGs is the possibility to evaluate both nEPPs and SPARQL with nEPPs queries under the RDFS entailment regime on existing query processors. We report on an experimental evaluation on a variety of real KGs.

Over the last decade, extensive research has been done on automatic construction of knowledge graphs from Web resources, resulting in a number of large-scale knowledge graphs such as YAGO, DBpedia, BabelNet, and Wikidata. Despite that some of these knowledge graphs are multilingual, they contain few or no linked data in Persian, and do not support tools for extracting knowledge from Persian information sources. FarsBase (available at http://farsbase.net/about) is the first Persian multi-source knowledge graph, which is specifically designed for semantic search engines to support Persian knowledge. FarsBase uses a diverse set of hybrid and flexible techniques to extract and integrate knowledge from various sources, such as Wikipedia, Web tables and unstructured texts. It also supports entity linking, which allows integration with other knowledge graphs. To maintain high accuracy for triples, we adopt a low-cost mechanism for verifying candidate knowledge by human experts, where the candidates for human verification are prioritized using different heuristics. FarsBase is being used as the semantic-search system of a Persian search engine and efficiently answers hundreds of semantic queries per second.