Effective use of BERT in graph embeddings for sparse knowledge graph completion

Liu, Xinglan; Hussain, Hussain; Razouk, Houssam; Kern, Roman

doi:10.1145/3477314.3507031

Cited by 4 publications

(8 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These knowledge bases attempt to model general commonsense knowledge including causal relationships. In context of common-sense knowledge completions, which involves predicting missing links in a knowledge base, text-aware knowledge graph embedding methods have shown to be effective [12], [13]. These methods leverage natural language processing techniques to extract information from the freeform text descriptions of entities in the knowledge base, which can then be used to improve the accuracy of the knowledge graph embedding.…”

Section: Related Workmentioning

confidence: 99%

“…Other approaches leverages language models to extract the text attributes embeddings. These embeddings are concatenated to nodes embeddings, which are learned for the knowledge graph structure [12] or even replaced them completely taking BERT-ConvE as a prime example [13].…”

Section: B Text Aware Knowledge Graph Emebdding For Common-sense Know...mentioning

confidence: 99%

“…At the same time, noteworthy advances in the commonsense knowledge completion techniques include new publicly available large scale data sets such as Atomic [10] and ConceptNet [11], modelling the common-sense knowledge bases as knowledge graphs and the use of knowledge graph embedding algorithms for downstream tasks such as common-sense knowledge graph completion [12], [13]. These advances have made it possible to integrate knowledge from various data sources.…”

Section: Introductionmentioning

confidence: 99%

“…These advances have made it possible to integrate knowledge from various data sources. For instance, in [12], [13] the authors leverage language models pre-trained on a large scale text data set to enrich the embedding of the nodes in common-sense knowledge bases. This technique can be especially important for root cause analysis and risk assessment in an industrial setting where a significant amount of knowledge is documented in a semi-structured way, as seen in FMEA, and in a unstructured textual format, as evident in failure analysis reports.…”

Section: Introductionmentioning

confidence: 99%

“…Complementary to existing research, that has explored ontology-based solutions for improving the transferability of causal domain knowledge contained in FMEA documents as outlined in [14]- [16] and has showcased notable progress in common-sense knowledge completion techniques [12], [13], our study aims to examine the potential impact of these advancements on the FMEA process effectiveness. Specifically, we explore whether leveraging common-sense knowledge completion techniques can enhance the comprehensibility of brainstorming sessions by treating FMEA documents as the initial knowledge base.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Improving FMEA Comprehensibility via Common-Sense Knowledge Graph Completion Techniques

Razouk,

Liu,

Kern

2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

The Failure Mode Effect Analysis process (FMEA) is widely used in industry for risk assessment, as it effectively captures and documents domain-specific knowledge. This process is mainly concerned with causal domain knowledge. In practical applications, FMEAs encounter challenges in terms of comprehensibility, particularly related to inadequate coverage of listed failure modes and their corresponding effects and causes. This can be attributed to the limitations of traditional brainstorming approaches typically employed in the FMEA process. Depending on the size and diversity in terms of disciplines of the team conducting the analysis, these approaches may not adequately capture a comprehensive range of failure modes, leading to gaps in coverage. To this end, methods for improving FMEA knowledge comprehensibility are highly needed. A potential approach to address this gap is rooted in recent advances in commonsense knowledge graph completion have illustrated effectiveness of text-aware graph embedding methods. However, the applicability of such methods in an industrial setting is limited. This paper addresses this issue on FMEA documents in an industrial environment. Here, the application of common-sense knowledge graph completion methods on FMEA documents from semiconductor manufacturing is studied. These methods achieve over 20% MRR on the test set and 70% of the top 10 predictions were manually assessed to be plausible by domain experts. Based on the evaluation, this paper confirms that text-aware knowledge graph embedding for common-sense knowledge graph completion are more effective than structure-only knowledge graph embedding for improving FMEA knowledge comprehensibility. Additionally we found that language model in domain fine-tuning is beneficial for extracting more meaningful embedding, thus improving the overall model performance.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: B Text Aware Knowledge Graph Emebdding For Common-sense Know...mentioning

confidence: 99%