SAKA: An Intelligent Platform for Semi-automated Knowledge Graph Construction and Application
Hanrong Zhang, Xinyue Wang, Jiabao Pan, Hongwei Wang
TL;DR
This paper tackles the usability gap in knowledge graph platforms and the underutilization of user-generated KGs, especially when converting audio data. It introduces SAKA, an intelligent platform that supports semi-automatic KG construction from structured data, audio-based KG information extraction (AGIE), and a semantic parsing-based KBQA system built on user-created KGs. The approach combines ontology-driven structured-data KG construction with audio-driven extraction (VAD, SD GE2E, and MIE) and a KBQA pipeline that uses Cypher over a Neo4j-backed KG, plus data crawling for domain KBs. Key contributions include a user-friendly KG construction workflow, the AGIE method for audio-to-KG creation, a KBQA module tailored to the medical domain, and a demonstration of substantial KG scale ($ ext{entities}\, ext{≈}\,33{,}000$, $ ext{relations}\, ext{≈}\,230{,}000$) along with multiple evaluation results (e.g., VAD accuracy $=97.42\%$, SD $EER=10.58\%$). The work enables broader access to KG-powered analytics and applications, especially in health contexts, and lays groundwork for multi-domain KG management and query-answering directly over user-built KGs.
Abstract
Knowledge graph (KG) technology is extensively utilized in many areas, and many companies offer applications based on KG. Nonetheless, most KG platforms necessitate expertise and tremendous time and effort from users to construct KG records manually, which poses great difficulties for ordinary people. Additionally, audio data is abundant and holds valuable information, but it is challenging to transform it into a KG. What's more, the platforms usually do not leverage the full potential of the KGs constructed by users. In this paper, we propose an intelligent and user-friendly platform for Semi-automated KG Construction and Application (SAKA) to address the aforementioned problems. Primarily, users can semi-automatically construct KGs from structured data of numerous areas by interacting with the platform, based on which multi-versions of KG can be stored, viewed, managed, and updated. Moreover, we propose an Audio-based KG Information Extraction (AGIE) method to establish KGs from audio data. Lastly, the platform creates a semantic parsing-based knowledge base question answering (KBQA) system based on the user-created KGs. We prove the feasibility of the semi-automatic KG construction method on the SAKA platform.