Computer Science > Artificial Intelligence
[Submitted on 20 Nov 2024]
Title:KAAE: Numerical Reasoning for Knowledge Graphs via Knowledge-aware Attributes Learning
View PDF HTML (experimental)Abstract:Numerical reasoning is pivotal in various artificial intelligence applications, such as natural language processing and recommender systems, where it involves using entities, relations, and attribute values (e.g., weight, length) to infer new factual relations (e.g., the Nile is longer than the Amazon). However, existing approaches encounter two critical challenges in modeling: (1) semantic relevance-the challenge of insufficiently capturing the necessary contextual interactions among entities, relations, and numerical attributes, often resulting in suboptimal inference; and (2) semantic ambiguity-the difficulty in accurately distinguishing ordinal relationships during numerical reasoning, which compromises the generation of high-quality samples and limits the effectiveness of contrastive learning. To address these challenges, we propose the novel Knowledge-Aware Attributes Embedding model (KAAE) for knowledge graph embeddings in numerical reasoning. Specifically, to overcome the challenge of semantic relevance, we introduce a Mixture-of-Experts-Knowledge-Aware (MoEKA) Encoder, designed to integrate the semantics of entities, relations, and numerical attributes into a joint semantic space. To tackle semantic ambiguity, we implement a new ordinal knowledge contrastive learning (OKCL) strategy that generates high-quality ordinal samples from the original data with the aid of ordinal relations, capturing fine-grained semantic nuances essential for accurate numerical reasoning. Experiments on three public benchmark datasets demonstrate the superior performance of KAAE across various attribute value distributions.
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.