Natural Language Query for Technical Knowledge Graph Navigation

Research output: Chapter in Book/Conference paperConference paperpeer-review


Technical knowledge graphs are difficult to navigate. To support users with no coding experience, one can use traditional structured HTML form controls, such as drop-down lists and check-boxes, to construct queries. However, this requires multiple clicks and selections. Natural language queries, on the other hand, are more convenient and less restrictive for knowledge graphs navigation. In this paper, we propose a system that enables natural language queries against technical knowledge graphs. Given an input utterance (i.e., a query in human language), we first perform Named Entity Recognition (NER) to identify domain specific entity mentions as node names, entity types as node labels, and question words (e.g., what, how many and list) as keywords of a structured query language before the rule-based formal query constructions. Three rules are exploited to generate a valid structured formal query. The web-based interactive application is developed to help maintainers access industrial maintenance knowledge graph which is constructed from text data.
Original languageEnglish
Title of host publicationData Mining - 20th Australasian Conference, AusDM 2022, Proceedings
EditorsLaurence A.F. Park, Simeon Simoff, Heitor Murilo Gomes, Maryam Doborjeh, Yee Ling Boo, Yun Sing Koh, Yanchang Zhao, Graham Williams
Place of PublicationSingapore
Number of pages16
ISBN (Electronic)9789811987465
ISBN (Print)9789811987458
Publication statusPublished - 2022
Event20th Australasian Data Mining Conference, AusDM 2022 - Western Sydney, Australia
Duration: 12 Dec 202215 Dec 2022

Publication series

NameCommunications in Computer and Information Science
Volume1741 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937


Conference20th Australasian Data Mining Conference, AusDM 2022
CityWestern Sydney


Dive into the research topics of 'Natural Language Query for Technical Knowledge Graph Navigation'. Together they form a unique fingerprint.

Cite this