Graph-based Knowledge Tracing: Modeling Student Proficiency Using Graph Neural Network

Graph-based Deep Q Network for Web Navigation

Generic graph (analysis) library in Typescript

A pure-Python structured memory benchmark for multi-agent LLM systems — context graph vs vector RAG vs raw history dump, five scenarios, 18 graded queries, zero API calls.

A Study of Graph-based System for Multi-view Clustering

GraphOfDocs: Representing multiple documents as a single graph

Graph-based Person Signature for Person Re-Identifications (CVPRW 21)

[ACM MM 2023] Scalable Incomplete Multi-View Clustering with Structure Alignment (SIMVC-SA)

Source code of partial-order hypergraph and baselines in our WWW 2018 paper, Learning on Partial-Order Hypergraphs

Modelling and analyzing the Covid19 Open Research using the GraphOfDocs model.

[T-NNLS] Unpaired Multi-View Graph Clustering with Cross-View Structure Matching

Website of TextGraphs: Graph-based Algorithms for Natural Language Processing

[ACM MM 2023] Efficient Multi-View Graph Clustering with Local and Global Structure Preservation

LangGraph is a powerful framework built on LangChain that enables the creation of stateful, multi-step, and agentic workflows using directed graphs. It simplifies complex LLM orchestration by allowing conditional branching, memory, and tool integrations in a visual and modular way.

The SSWG/MVM framework is a recursive workflow generator that unites symbolic structure with semantic evaluation. It synthesizes workflows, validates them against schemas, measures clarity and coherence, and iteratively refines results until improvement plateaus. Through memory, metrics, and modular orchestration, it is both software and mindset.

An implementation of paper "Expectation-Guided Self-Verification for Aligning Large Reasoning Models with Domain Knowledge"

A prototype system for transforming raw thought streams into a structured, queryable idea graph using LLMs and embeddings.

Topological Graph Memory for Reinforcement Learning — 100% retention, zero hidden layers. Replaces neural weights with navigable graph topology to eliminate catastrophic forgetting.

Python/R methodological reproduction of Seurat v4 Weighted Nearest Neighbor analysis for multimodal single-cell RNA and ADT data integration.

Datasets and evaluation results for "An innovative graph-based approach to advance feature selection from multiple textual documents" paper