Pular para o conteúdo principal

Knowledge Graph Toolkit (KGTK)

https://kgtk.readthedocs.io/en/latest/

Diagram

KGTK represents KGs using TSV files with 4 columns labeled id, node1, label and node2. The id column is a symbol representing an identifier of an edge, corresponding to the orange circles in the diagram above. node1 represents the source of the edge, node2 represents the destination of the edge, and label represents the relation between node1 and node2.

>> Quad do RDF, definir cada tripla como um grafo

Diagram>< 

KGTK defines knowledge graphs (or more generally any attributed graph or hypergraph) as a set of nodes and a set of edges between those nodes. KGTK represents everything of meaning via an edge. Edges themselves can be attributed by having edges asserted about them, thus, KGTK can in fact represent arbitrary hypergraphs. KGTK intentionally does not distinguish attributes or qualifiers on nodes and edges from full-fledged edges, tools operating on KGTK graphs can instead interpret edges differently if they so desire. In KGTK, everything can be a node, and every node can have any type of edge to any other node.


import-ntriples: This command will import one or more ntriple files into KGTK format.

>> Existem comando de importação para KGs como Wikidata

The generate-wikidata-triples command generates triple files from a kgtk file. The generated triple files can then be loaded into a triple store directly.

The triple generator reads a tab-separated kgtk file from standard input, by default, or a given file. The kgtk file is required to have at least the following 4 fields: node1, label, node2 and id. The node1 field is the subject; label is the predicate and node2 is the object. 

>> Converter Wikidata em RDF para carregar em TripleStore

Transformation commands
calc: this command performs calculations on one or more columns in a KGTK file. The output of a calculation can be written into an existing column or into a new column, which will be added after all existing columns.

lexicalize builds English sentences from KGTK edge files. The primary purpose of this command is to construct inputs for text-based distance vector analysis. However, it may also prove useful for explaining the contents of local subsets of Knowledge Graphs.

Curation commands
validate-properties validates and filter property patterns in a KGTK file. We want to be able to detect violations of various constraint patterns.
An existing system, SHACL, is an RDF-based constraint system. We'd like KGTK to have something that is both easier for new users than RDF and more efficient to run.

Analysis
find the connected components in a KGTK edge file.
compute the the embeddings of this files' entities. We are using structure of nodes and their relations to compute embeddings of nodes. The set of metrics to compute are specified by the user. There are three supported formats: glove, w2v, and kgtk. The algorithm by default is ComplEx (also supports TransE, DistMult, or RESCAL)
compute centrality metrics and connectivity statistics.
computes paths between each pair of source-target nodes
find all nodes reachable from given root nodes in a KGTK edge file. That is, given a set of nodes N and a set of properties P, this command computes the set of nodes R that can be reached from N via paths containing any of the properties in P.
Computes embeddings of nodes using properties of nodes. The values are concatenated into sentences defined by a template, and embedded using a pre-trained language model.

Artigo -> https://arxiv.org/pdf/2006.00088.pdf


Comentários

Postar um comentário

Sinta-se a vontade para comentar. Críticas construtivas são sempre bem vindas.

Postagens mais visitadas deste blog

Connected Papers: Uma abordagem alternativa para revisão da literatura

Durante um projeto de pesquisa podemos encontrar um artigo que nos identificamos em termos de problema de pesquisa e também de solução. Então surge a vontade de saber como essa área de pesquisa se desenvolveu até chegar a esse ponto ou quais desdobramentos ocorreram a partir dessa solução proposta para identificar o estado da arte nesse tema. Podemos seguir duas abordagens:  realizar uma revisão sistemática usando palavras chaves que melhor caracterizam o tema em bibliotecas digitais de referência para encontrar artigos relacionados ou realizar snowballing ancorado nesse artigo que identificamos previamente, explorando os artigos citados (backward) ou os artigos que o citam (forward)  Mas a ferramenta Connected Papers propõe uma abordagem alternativa para essa busca. O problema inicial é dado um artigo de interesse, precisamos encontrar outros artigos relacionados de "certa forma". Find different methods and approaches to the same subject Track down the state of the art rese...

Knowledge Graph Embedding with Triple Context - Leitura de Abstract

  Jun Shi, Huan Gao, Guilin Qi, and Zhangquan Zhou. 2017. Knowledge Graph Embedding with Triple Context. In Proceedings of the 2017 ACM on Conference on Information and Knowledge Management (CIKM '17). Association for Computing Machinery, New York, NY, USA, 2299–2302. https://doi.org/10.1145/3132847.3133119 ABSTRACT Knowledge graph embedding, which aims to represent entities and relations in vector spaces, has shown outstanding performance on a few knowledge graph completion tasks. Most existing methods are based on the assumption that a knowledge graph is a set of separate triples, ignoring rich graph features, i.e., structural information in the graph. In this paper, we take advantages of structures in knowledge graphs, especially local structures around a triple, which we refer to as triple context. We then propose a Triple-Context-based knowledge Embedding model (TCE). For each triple, two kinds of structure information are considered as its context in the graph; one is the out...

KnOD 2021

Beyond Facts: Online Discourse and Knowledge Graphs A preface to the proceedings of the 1st International Workshop on Knowledge Graphs for Online Discourse Analysis (KnOD 2021, co-located with TheWebConf’21) https://ceur-ws.org/Vol-2877/preface.pdf https://knod2021.wordpress.com/   ABSTRACT Expressing opinions and interacting with others on the Web has led to the production of an abundance of online discourse data, such as claims and viewpoints on controversial topics, their sources and contexts . This data constitutes a valuable source of insights for studies into misinformation spread, bias reinforcement, echo chambers or political agenda setting. While knowledge graphs promise to provide the key to a Web of structured information, they are mainly focused on facts without keeping track of the diversity, connection or temporal evolution of online discourse data. As opposed to facts, claims are inherently more complex. Their interpretation strongly depends on the context and a vari...