NER-Luxury: Named entity recognition for the fashion and luxury domain
Paper
•
2409.15804
•
Published
You need to agree to share your contact information to access this model
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
You agree to not use the model to conduct experiments that cause harm to human subjects.
Log in or Sign Up to review the conditions and access this model content.
NER-Luxury
A fine-tuned XLM-Roberta model for NER in the fashion and luxury industry
. Goal
- NER-Luxury is a fine-tuned XLM-Roberta model for the subtask N.E.R (Named Entity Recognition) in English. NER-Luxury is domain-specific for the fashion and luxury industry with bespoke labels. NER-Luxury is trying to be a bridge between the aesthetic side and the quantitative side of the fashion and luxury industry.
- As a downstream task, NER-Luxury is able to identify major fashion houses, artistic directors, fragrances, models, or influential artists on the website of a fashion magazine. And NER-Luxury is also able to identify companies, listed groups, executives, financial analysts, and investment companies inside a 200-page quarterly financial report.
- The goal of NER-Luxury is to create a clear hierarchical classification of luxury houses, fine watchmakers, beauty brands, sportswear labels, and fast fashion brands with respect of temporality, context, and sustainability. NER-Luxury is trying to solve the "entity disambiguation" between the founder, his eponymous label, the company designation, the names of products, and the intellectual property rights for corporate lawyers, M&A bankers, and financial analysts.
For example, the disambiguation of Louis Vuitton:
- The visionary founder, Louis Vuitton (1821-1892)
- The luxury house, Louis Vuitton
- The giant luxury group LVMH Moët Hennessy Louis Vuitton SE
- The collection with Japanese artist, Louis Vuitton x Yayoi Kusama
. NER bespoke labels
Entities are evolving according to temporality, and context.
| Label | Description and example |
|---|---|
| O | Outside (of a text segment) |
| Date | Temporal expressions (1854, Q2 2023, Nineties, September 21) |
| Location | Physical location and area (Paris, Japan, Europe, Champs-Elysées) |
| Event | Critical events (WW II, Olympics, IPO, Covid pandemic, Paris Fashion Week) |
| MonetaryValue | Currency, price, sales, revenue ($2.65 billion, 4.6 million euros, CHF 400,000, etc.) |
| House | Fashion and luxury houses (Louis Vuitton, Cartier, Gucci, Chanel) |
| Brand | Sportswear, beauty and labels (Nike, Lululemon, Clinique) |
| FastFashion | Mass-market retailers (Zara, H&M, Uniqlo, Shein) |
| PrivateCompany | Unlisted companies (Chanel SA, Stella McCartney Ltd, Valentino S.p.A) |
| ListedGroup | Listed groups (LVMH, Hermès International SCA, Kering) |
| HoldingTrust | Holding and family office (Agache, H51, Mousse Partners, Artèmis) |
| InvestmentFirm | Investment banks, PE funds, M&A firms (KKR, L Catterton, Mayhoola, Bernstein) |
| MediaPublisher | Media outlets (Bloomberg, Vogue, Business of Fashion, NYT) |
| Hospitality | Luxury hospitality (Ritz Paris, Belmond hotel Cipriani,Venetian Macao) |
| MuseumGallery | Exhibition spaces (Louvre, MET, Victoria & Albert, Pinault Collection) |
| Retailer | POS, department stores, and select shops (Bergdorf, Le Bon Marché, Takashimaya) |
| Education | Business and fashion schools (Polytechnic, Harvard, LSE, ESCP, Central Saint Martins, IFM) |
| Organization | Legal, scientific, and cultural entities (CFDA, European Union, UNESCO, SEC) |
| ArtisticDirector | Lead creative of houses (Karl Lagerfeld, Daniel Lee, Sarah Burton, Alessandro Michele) |
| Executive | C-level, board members (Jérôme Lambert, Sue Nabi, Pietro Beccari) |
| Founder | Founder, creative, and owner (Ralph Lauren, Rei Kawakubo, Michael Kors) |
| Chairperson | Chairman/Chairwoman (Bernard Arnault, Patrizio Bertelli, François-Henri Pinault) |
| AnalystBanker | Equity analysts, M&A bankers (Luca Solca, Pierre Mallevays, Louise Singlehurst) |
| KOL | Artists, celebrities, historical figures (Audrey Hepburn, BTS, Kanye West, Emma Watson) |
| AthleteTeam | Professional athletes and teams (David Beckham, Maria Sharapova, Luna Rossa, Scuderia Ferrari) |
| Model | Fashion models (Iman, Kate Moss, Adriana Lima, Naomi Campbell, Mariacarla Boscono) |
| CreativeInsider | Photographers, make-up artists, watchmakers (Steven Meisel, Dominique Ropion, Gérald Genta) |
| EditorJournalist | Editor-in-chief, fashion editors, journalists (Suzy Menkes, Anna Wintour, Carine Roitfeld) |
| GarmCollection | Iconic garment and collections (Haute Couture, Bar suit, No.13 of McQueen, Green Jungle Dress) |
| Cosmetic | Cosmetic products (Tilbury Glow palette, Crème de La Mer, YSL Nu, Viva Glam) |
| Fragrance | Perfumes and EdT (Chanel No.5, Dior Sauvage, Terre d'Hermès, Tom Ford Black Orchid) |
| BagTrvlGoods | Bags, handbags, and leather goods (Hermès Birkin bag, Louis Vuitton Speedy bag, Chanel 2.55) |
| Jewelry | Fine jewellery, and gems (Alhambra of Van Cleef & Arpels, Juste un Clou Cartier, The Winston Blue) |
| Timepiece | Fine watches (Nautilus Patek Philippe, Reverso Jaeger-Lecoultre, Rolex Oyster) |
| Footwear | High heels to sneakers (Rainbow of Ferragamo, Armadillo of McQueen, Air Force1) |
| WineSpirit | Wine and spirit (Château d'Yquem, Clos de Tart, Château Matras, Hennessy, Moet, Belvedere) |
| Sustainability | Relevant ESG factors and entities (Ethical Fashion Initiative, decoupling, biodiversity loss) |
| CulturalArtifact | Songs, books, movies (The Devil wears Prada, American Gigolo, Poker Face, The College Dropout) |
Paper address and cite information: https://arxiv.org/abs/2409.15804
Citation info
@misc{mousterou2024nerluxurynamedentityrecognition,
title={NER-Luxury: Named entity recognition for the fashion and luxury domain},
author={Akim Mousterou},
year={2024},
eprint={2409.15804},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2409.15804},
}
How to use NER-Luxury with HuggingFace?
Load NER-Luxury and its sub-word tokenizer :
from transformers import AutoTokenizer, AutoModelForTokenClassification
from transformers import pipeline
tokenizer = AutoTokenizer.from_pretrained("AkimfromParis/NER-Luxury")
model = AutoModelForTokenClassification.from_pretrained("AkimfromParis/NER-Luxury")
nlp = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="simple")
example = "CEO Leena Nair dismisses IPO rumours for Chanel."
ner_results = nlp(example)
print(ner_results)
NER-Luxury
This model is a fine-tuned version of xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4079
- Precision: 0.7652
- Recall: 0.8033
- F1: 0.7838
- Accuracy: 0.9403
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 8
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 1.1269 | 1.0 | 1155 | 0.6237 | 0.6085 | 0.6716 | 0.6385 | 0.9005 |
| 0.5871 | 2.0 | 2310 | 0.4933 | 0.6857 | 0.7367 | 0.7103 | 0.9208 |
| 0.4517 | 3.0 | 3465 | 0.4470 | 0.7115 | 0.7639 | 0.7368 | 0.9273 |
| 0.3692 | 4.0 | 4620 | 0.4271 | 0.7298 | 0.7797 | 0.7539 | 0.9322 |
| 0.3121 | 5.0 | 5775 | 0.4103 | 0.7422 | 0.7906 | 0.7656 | 0.9362 |
| 0.2726 | 6.0 | 6930 | 0.4109 | 0.7531 | 0.7940 | 0.7730 | 0.9381 |
| 0.2138 | 7.0 | 8085 | 0.4088 | 0.7632 | 0.8005 | 0.7814 | 0.9397 |
| 0.1962 | 8.0 | 9240 | 0.4079 | 0.7652 | 0.8033 | 0.7838 | 0.9403 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.0+cu121
- Datasets 2.21.0
- Tokenizers 0.19.1
- Downloads last month
- -
Model tree for AkimfromParis/NER-Luxury
Base model
FacebookAI/xlm-roberta-baseSpace using AkimfromParis/NER-Luxury 1
Paper for AkimfromParis/NER-Luxury
Evaluation results
- Loss on Privateself-reported0.408
- Precision on Privateself-reported0.765
- Recall on Privateself-reported0.803
- F1 on Privateself-reported0.784
- Accuracy on Privateself-reported0.940