ArticleScientific reports2026
Dynamic teaching path generation for quadruped robot programming by integrating R-GCN and SBERT.
Article in Scientific reports, 2026. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Not yet cited in PubMed.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
0 citing papers in PubMed.
No citing paper in PubMed yet.
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
1 author.
Funding
No grant is acknowledged in the PubMed record.
Abstract
To address the problems of knowledge circular dependency and learning path ambiguity in quadruped robot programming learning, a dynamic teaching path generation method that integrates structural and semantic information is proposed. A domain knowledge graph covering core knowledge and prerequisite relationships is constructed. R-GCN (Relational Graph Convolutional Network) is used to generate structured node embeddings through link prediction tasks. SBERT (Sentence-Bidirectional Encoder Representations from Transformers) is also used to generate semantic vector representations of knowledge points and user tasks. For a given programming task, seed nodes are retrieved based on semantic similarity, and a priori knowledge subgraph is constructed by back-tracing. The number of seed nodes is fixed to the top 5 knowledge points ranked by semantic similarity. The similarity threshold is implicitly determined by this ranking strategy rather than a fixed value. The weighting coefficients used in edge scoring and node prioritization are set to α = 0.6 and β = 0.7 based on validation experiments described in Sect. 2.4 and 2.5. All models are implemented using PyTorch and trained on a single NVIDIA RTX 3090 GPU. The batch size is set to 512, and early stopping is applied based on validation loss with a patience of 10 epochs. Relationship strength is quantified by calculating the dot product of the embedding vectors, automatically identifying and pruning the weakest links in circular dependencies to achieve intelligent link breaking. During topological sorting, nodes with high semantic match to the task are prioritized for learning path generation. Experiments show that the proposed method achieves superior performance in the quadruped robot programming task, with a Top-3 score of 0.88, higher than CompGCN's 0.73, a mean reciprocal rank of 0.68, and path integrity and structural rationality reaching 0.85 and 0.88, respectively. This method performs well in knowledge association modeling and path recommendation, effectively supports structured, task-driven programming teaching, and provides cognitively coherent learning support for beginners.
Indexed as
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.