SemSur: A Core Ontology for the Semantic Representation of Research Findings

Fathalla, Said; Vahdati, Sahar; Auer, Sören; Lange, Christoph

doi:10.1016/j.procs.2018.09.015

Cited by 20 publications

(9 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, based on the dataset, we would expect the response to CQ1 to be that LLMs (which are sub-types of GenAI) exhibit hallucination, and the response to CQ2 to be the Factuality hallucination type is the most frequent type to be present. To conduct the actual evaluation, as is standard in other related work, 46,47 we first translate the CQs (expressed informally in natural language as shown in Table 3) into formal SPARQL queries framed using the concepts and properties in HALO. We note that, because HALO contains both a Hallucination and Metadata module, it contains the necessary ontological elements for modeling the CQs as equivalent SPARQL queries.…”

Section: Ontology Evaluationmentioning

confidence: 99%

HALO: an ontology for representing and categorizing hallucinations in large language models

Nananukul,

Kejriwal

2024

Disruptive Technologies in Information Sciences VIII

View full text Add to dashboard Cite

Recent progress in generative AI, including large language models (LLMs) like ChatGPT, has opened up significant opportunities in fields ranging from natural language processing to knowledge discovery and data mining. However, there is also a growing awareness that the models can be prone to problems such as making information up or 'hallucinations', and faulty reasoning on seemingly simple problems. Because of the popularity of models like ChatGPT, both academic scholars and citizen scientists have documented hallucinations of several different types and severity. Despite this body of work, a formal model for describing and representing these hallucinations (with relevant meta-data) at a fine-grained level, is still lacking. In this paper, we address this gap by presenting the Hallucination Ontology or HALO, a formal, extensible ontology written in OWL that currently offers support for six different types of hallucinations known to arise in LLMs, along with support for provenance and experimental metadata. We also collect and publish a dataset containing hallucinations that we inductively gathered across multiple independent Web sources, and show that HALO can be successfully used to model this dataset and answer competency questions.

show abstract

Section: Ontology Evaluationmentioning

confidence: 99%

HALO: an ontology for representing and categorizing hallucinations in large language models

Nananukul,

Kejriwal

2024

Disruptive Technologies in Information Sciences VIII

View full text Add to dashboard Cite

show abstract

“…OpenResearch.org does not only represent scholarly events but also the scientific contributions within publications (Fathalla et al. 2017b , 2018 ).…”

Section: Related Workmentioning

confidence: 99%

A comprehensive quality assessment framework for scientific events

et al. 2020

Self Cite

View full text Add to dashboard Cite

Systematic assessment of scientific events has become increasingly important for research communities. A range of metrics (e.g., citations, h-index) have been developed by different research communities to make such assessments effectual. However, most of the metrics for assessing the quality of less formal publication venues and events have not yet deeply investigated. It is also rather challenging to develop respective metrics because each research community has its own formal and informal rules of communication and quality standards. In this article, we develop a comprehensive framework of assessment metrics for evaluating scientific events and involved stakeholders. The resulting quality metrics are determined with respect to three general categories—events, persons, and bibliometrics. Our assessment methodology is empirically applied to several series of computer science events, such as conferences and workshops, using publicly available data for determining quality metrics. We show that the metrics’ values coincide with the intuitive agreement of the community on its “top conferences”. Our results demonstrate that highly-ranked events share similar profiles, including the provision of outstanding reviews, visiting diverse locations, having reputed people involved, and renowned sponsors.

show abstract

“…Core ontologies provide a precise definition of structural knowledge in a specific field that connects different application domains [3,4,5,11]. They are located in the layer between upper-level (fundamental) and domain-specific ontologies, providing the definition of the core concepts from a specific field.…”

Section: Related Workmentioning

confidence: 99%

“…They aim at linking general concepts of a top-level ontology to more domain-specific concepts from a sub-field. Even though there is a large body of work making use of ontologies as a formal basis to model different aspects of scientific research, such as [3,4,5], few studies have focused on modeling scientific hypotheses [2,6].…”

Section: Related Workmentioning

confidence: 99%

Towards a Core Ontology for Hierarchies of Hypotheses in Invasion Biology

Algergawy¹,

Stangneth²,

Heger

et al. 2020

The Semantic Web: ESWC 2020 Satellite Events

View full text Add to dashboard Cite

With a rapidly growing body of knowledge, it becomes more and more difficult to keep track of the state of the art in a research field. A formal representation of the hypotheses in the field, their relations, the studies that support or question them based on which evidence, would greatly ease this task and help direct future research efforts. We present the design of such a core ontology for one specific field, namely invasion biology. We introduce the design of the Hierarchy of Hypotheses (HoH) core ontology to semantically capture and model the information contained in hierarchies of hypotheses created for invasion biology. The proposed core ontology is based on a number of well structured related ontologies, which provide a solid basis for its main entities.

show abstract

SemSur: A Core Ontology for the Semantic Representation of Research Findings

Cited by 20 publications

References 12 publications

HALO: an ontology for representing and categorizing hallucinations in large language models

HALO: an ontology for representing and categorizing hallucinations in large language models

A comprehensive quality assessment framework for scientific events

Towards a Core Ontology for Hierarchies of Hypotheses in Invasion Biology

Contact Info

Product

Resources

About