The scarcity of fully-annotated data becomes the biggest obstacle that prevents many deep learning approaches from widely applied. Weakly-supervised visual which can utilize inexact annotations is developed rapidly to remedy such a situation. In this paper, we study weakly-supervised task achieving pixel-level semantic segmentation only with image-level labels as supervision. Different other me...

The Natural History Museum, London (NHM), in collaboration with Amazon Web Services (AWS), has embarked on a project to build the Planetary Knowledge Base (PKB), comprehensive graph network comprising data all specimens, collectors, and localities. In initial prototype, we have concentrated botanical using plant taxa specimens within Global Biodiversity Information Facility (GBIF), combined geo...

A key problem in semantic parsing with graph-based semantic representations is graph parsing, i.e. computing all possible analyses of a given graph according to a grammar. This problem arises in training synchronous string-to-graph grammars, and when generating strings from them. We present two algorithms for graph parsing (bottom-up and top-down) with s-graph grammars. On the related problem o...

Finding Semantic Web statements The Semantic Web can be seen as a large knowledge-base of statements about resources, forming an interconnected graph through multiple references to the same resources. But the graphs in the Semantic Web are decentralised: there is not one single knowledge base that contains the graph of statements but instead anyone can contribute statements on his “personal” we...

The continuing development of Semantic Web technologies and the increasing user adoption in the recent years have accelerated the progress incorporating explicit semantics with data on the Web. With the rapidly growing RDF (Resource Description Framework) data on the Semantic Web, processing large semantic graph data have become more challenging. Constructing a summary graph structure from the ...

In this paper, we present a methodology, called Semantic Graph Mining, for computer-aided extraction of actionable rules from consolidated semantic graphs of statements. First, generate semantic annotations of a set of heterogeneous knowledge/information resources in terms of domain ontology. Second, merge a semantic graph by means of semantic integration of the annotated resources. Third, disc...

In this work we propose an algorithm for automatic speech understanding that takes a word graph as its input. First, this word graph is processed by means of a dynamic programming algorithm which gives as a result a second graph that includes semantic information. Computing the best path over this second graph allows us to obtain the most likely concept sequence, given the acoustic evidence ref...