A Content Based E-Commerce Dataset Recommendation System Using BERT and Named Entity Recognition
Bibliographic record
Abstract
Existing dataset recommendation (rec) systems including those named as ZhangRec23, WangRec22, and GDS19, face certain limitations, such as lack of focus on e-commerce datasets, inability to address complex queries, and reliance on inconsistent metadata (e.g., data structure of domain of products being recommended). This leads to incomplete or mismatched results returned by the system for complex query searches, such as "impact of seasonal sales on customer reviews for electronics". These traditional dataset rec systems rely on simple keyword matching, failing to interpret context-sensitive queries that researchers often need, and are unable to capture the dynamic trends in the e-commerce domain. This highlights the need for an advanced dataset rec system that improves metadata quality and integrates semantic understanding to recommend precise and relevant e-commerce datasets to researchers. This paper proposes an E-commerce Datasets Mining Rec System (EDMRec), an adaptation of ZhangRec23 approach. EDMRec combines content-based filtering, advanced metadata processing, and machine learning approach in a three-layered structure involving (i) Data Collection, (ii)Data Processing, and (iii) Query Processing. It utilizes Named Entity Recognition (NER) to complete metadata and uses TF-IDF with Bidirectional Encoder Representations from Transformers (BERT) embeddings to capture both keyword relevance and semantic context, enhancing recommendation precision for complex queries. Experimental results show that EDMRec improves precision, recall, and F1 score by 15% over existing systems, consistently providing contextually accurate recommendations across 4,373 metadata entries from sources such as Kaggle and Google Dataset Search, making it well-suited for supporting data-driven insights in e-commerce.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.006 | 0.006 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.004 |
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".