Special Issue on "Foundation Models for Information Retrieval and Knowledge Processing"

Motivation and Scope

Foundation models are pre-trained based on massive amounts of data and serve as the base for various downstream tasks. The most representative one is the Large Language Model ChatGPT, that has completely transformed the landscape of natural language processing (NLP), delivering remarkable performance for text classification/summarization, named entity recognition, translation, sentiment analysis, and so on. Pre-trained computer vision (CV) models such as VGG-16 and ResNet50 based on ImageNet have also been used as foundation models for various vision tasks including object detection, image classification, or semantic segmentation. Inspired by the success of these CV/NLP foundation models, this proposal intends to bring together researchers to work on building foundation models for information retrieval and knowledge processing, to enhance the efficiency, accuracy, and understanding of the information retrieval and knowledge processing systems and models, ultimately empowering users to access and leverage vast amounts of information effectively. Specifically, we expect to achieve enhanced accuracy in information retrieval tasks and neural-based forecasting models, efficient extraction and understanding of large multimodal data sources, and superior adaptability for various industry-specific or domain-specific scenarios.

Topics of interest include, but are not limited to the following:

• Novel datasets and benchmarks for building domain-specific foundation models
• Foundation models for recommendation
• Foundation models for ranking and prediction
• Foundation models for knowledge representation and reasoning
• Multimodal data analysis with foundation models
• Large-scale data processing with foundation models
• Pre-training techniques for foundation models
• Fine-tuning techniques for foundation models
• Prompt engineering techniques for foundation models
• Transfer learning techniques for foundation models
• Domain adaptation techniques for foundation models
• Domain-specific foundation models (e.g., spatiotemporal, social media, heath, knowledge graph)
• Systems and applications based on foundation models
• New applications based on knowledge graph, information retrieval, and foundation models

Guest Editors:

Shuo Shang ([email protected]), Professor, University of Electronic Science and Technology of China, China

Renhe Jiang ([email protected]), Assistant Professor, University of Tokyo, Japan

Ryosuke Shibasaki ([email protected]), Professor, University of Tokyo, Japan

Rui Yan ([email protected]), Associate Professor, Renmin University of China, China

Important Dates:

• Submission portal opens: November 15, 2023
• Deadline for paper submission: March 31, 2024
• First notification: May 31, 2024
• Revision: July 31, 2024
• Final decision: August 31, 2024

How to submit:

Authors interested in this special issue should submit their work to the manuscript tracking system at https://mc03.manuscriptcentral.com/di. Please choose “the special issue on FMIRKP” when submitting or indicate in the cover letter that you are submitting to the "Special Issue on FMIRKP”".

For further inquiries about submission, please contact the managing editor Dr. Fenghong Liu at [email protected]