Competitive Analysis with Graph Embedding on Patent Networks
Bibliographic record
Abstract
Advanced competitive analysis is increasingly becoming important in business analytics. A key component of strategic research is to collect and review information from multiple unstructured sources to identify major competitors and their technology development trends. Topic modeling techniques such as Latent Dirichlet Allocation (LDA) have been applied to competitive analysis, which mainly use the semantic similarities between documents to infer competitive relationships. In this study, we propose using graph embedding methods to learn the implicit competitive relationships between firms. Using patent networks with patents and organizations as nodes, we learn the embeddings of nodes which are then used to cluster organizations into groups. Organizations within the same groups are considered competitors. We applied three graph embedding methods: node2vec, metapath2vec, and GraphSAGE to learn node embeddings. Two of these methods use the structural information in patent networks: node2vec for homogeneous networks and metapath2vec for heterogeneous networks. While, GraphSAGE uses both the structure and content information in the patent network. The results are compared with a baseline author-topic modeling method. The graph embedding methods outperform the author-topic modeling approach in learning the competitive relationships. A case study, examining the evolution of competitors over multiple years, shows the graph embedding method learns meaningful node embeddings.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.004 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".