隨著電商市場的蓬勃發展，產品的日新月異下，會不斷的出現有用的新特徵類別和特徵值。不過，在還沒有探討新特徵類別的問題前，就已經有遺失值問題，如果平台定義更多的類別，要求賣家輸入對應的值，是不太可行的。又有些新特徵類別，連電商平台廠商都無法完整定義給賣家。因此有進階的遺失特徵問題。開放世界特徵探勘(Open-World Attribute Mining, OAM)的研究，雖能緩解遺失特徵的影響，但還無法自動命名新特徵類別。本研究定義了完全開放特徵探勘(Fully Open Attribute Mining, FOAM)任務，以完全開放的方式自動探勘文本上的所有特徵，來完整解決遺失特徵問題。FOAM在解決電商平台遺失特徵下，協助賣家上架，也能連帶讓一些電商平台和買家也能受惠於產品搜尋及推薦系統的提升。
本研究提出了FOA-Mine來直接達成FOAM。對於候選特徵值擷取，會利用詞類(Part of Speech, POS)正規表達式(Regex)學習，全自動地找出通用POS Regex，來擷取出品質更高的候選特徵值。此外，也會利用搭配自動優化種子閾δ的已知類別監督式分類來探勘候選特徵值，並與大型語言模型(Large Language Model, LLM)的探勘結果集成，來產生最終高品質的特徵探勘結果。LLM除了確保最終所有特徵值都有被命名的特徵類別，LLM探勘中全新的明確格式Prompt組件與已知類別監督式分類的搭配，也讓FOA-Mine能在捨棄RAG架構下，再提升LLM探勘的結果。
在利用Amazon產品資料做實驗後，FOA-Mine裡的POS Regex擷取除了比現有SOTA的POS候選特徵值擷取多了2.5%的F1，本研究也發現候選特徵值和LLM能互相提升最終的探勘品質。此外，FOA-Mine在FOAM的綜合評分fARB上也能比現有SOTA高2.8%。而本研究的FOA-Mine和LLM探勘(簡稱FLLM)在未知類別探勘的綜合評分rARB上，也都能比現有SOTA多1倍左右的分數。本研究的成果，除了對電商有意義，也期望能協助到許多非電商領域的研究。

The booming e-commerce market, with its constantly evolving product features, has led to a growing challenge: advanced missing attribute problems. This paper defines the Fully Open Attribute Mining (FOAM) task, which addresses the shortcomings of Open-World Attribute Mining by mining all attributes from text in a fully automation way, thereby providing a comprehensive solution to the missing attribute problem. FOAM not only solves the missing attributes of e-commerce platforms, but also assists sellers in listing their products, and can also benefit some e-commerce platforms and buyers through the improvement of product search and recommendation systems.
This research proposes FOA-Mine to accomplish the FOAM task. For candidate attribute value extraction, FOA-Mine automatically learns generalized Part of Speech (POS) Regular Expressions (Regex) through POS sequence filtering, enabling the extraction of higher-quality candidate attribute values. Furthermore, the study utilizes known-type supervised classification with automatically optimized seed thresholds to mine attribute types of candidate attribute values. Then the supervised result is ensembled with Large Language Model (LLM) mining results. The LLM mining will also use novel explicit format prompt component to enhance mining quality. After the ensemble, it allows FOA-Mine to significantly improve LLM's mining performance even without a Retrieval Augmented Generation (RAG) architecture, and also ensuring that all attribute values are assigned with semantic attribute types.
Experimental results on the Amazon product dataset demonstrate that FOA-Mine's POS Regex extraction improves the F1 score by 2.5% compared to existing state-of-the-art (SOTA) POS candidate attribute value extraction methods. The study also reveals that candidate attribute values and the LLM mutually enhance the final mining quality. Furthermore, FOA-Mine's comprehensive FOAM score, fARB, is 2.8% higher than SOTA. For unknown-type attribute mining, FOA-Mine achieves approximately double the score of SOTA on the comprehensive metric rARB.

Abderzag, F. (2021). Markets and the creation of competitive advantages in companies according to an internal marketing orientation. International Journal of ADVANCED AND APPLIED SCIENCES, 8, 93-107. https://doi.org/10.21833/ijaas.2021.10.011
Adnabu. (2024). 65 Amazon Seller Statistics You Must Know for 2024. Adnabu. Retrieved 16 October 2024 from https://blog.adnabu.com/amazon/amazon-seller-statistics/
Agarwal, R., Singh, A., Zhang, L., Bohnet, B., Rosias, L., Chan, S., Zhang, B., Anand, A., Abbas, Z., & Nova, A. (2024). Many-shot in-context learning. Advances in neural information processing systems, 37, 76930-76966.
Akbik, A., Bergmann, T., Blythe, D., Rasul, K., Schweter, S., & Vollgraf, R. (2019, June). FLAIR: An Easy-to-Use Framework for State-of-the-Art NLP. In W. Ammar, A. Louis, & N. Mostafazadeh, Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics (Demonstrations) Minneapolis, Minnesota.
Alibrahim, H., & Ludwig, S. A. (2021, 28 June-1 July 2021). Hyperparameter Optimization: Comparing Genetic Algorithm against Grid Search and Bayesian Optimization. 2021 IEEE Congress on Evolutionary Computation (CEC),
Bae, S., Kim, S., Ko, J., Lee, G., Noh, S., & Yun, S.-Y. (2023). Self-Contrastive Learning: Single-Viewed Supervised Contrastive Framework Using Sub-network. Proceedings of the AAAI Conference on Artificial Intelligence, 37(1), 197-205. https://doi.org/10.1609/aaai.v37i1.25091
Bird, Klein, S. a., Loper, E. a., & Edward. (2009). Natural Language Processing with Python: Analyzing Text with the Natural Language Toolkit. O'Reilly Media, Inc. https://www.nltk.org/book/
Cao, X., Liu, Y., Hu, B., & Zhang, Y. (2021). Dual‐Channel Reasoning Model for Complex Question Answering. Complexity, 2021(1), 7367181.
Cao, Y., Geddes, T. A., Yang, J. Y. H., & Yang, P. (2020). Ensemble deep learning in bioinformatics. Nature Machine Intelligence, 2(9), 500-508. https://doi.org/10.1038/s42256-020-0217-y
Chen, B., Zhang, Z., Langrené, N., & Zhu, S. (2023). Unleashing the potential of prompt engineering in large language models: a comprehensive review. arXiv preprint arXiv:2310.14735.
Davenport, L. (2020). How To Create Amazon Product Listings With Variations | Step-by-Step. Youtube. Retrieved 24 October 2024 from https://www.youtube.com/watch?v=ferFzzCW4ls
De la Comble, A., Dutt, A., Montalvo, P., & Salah, A. (2022). Multi-modal attribute extraction for e-commerce. arXiv preprint arXiv:2203.03441.
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018). Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805.
Fan, W., Ding, Y., Ning, L., Wang, S., Li, H., Yin, D., Chua, T.-S., & Li, Q. (2024). A Survey on RAG Meeting LLMs: Towards Retrieval-Augmented Large Language Models Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, Barcelona, Spain. https://doi.org/10.1145/3637528.3671470
Fang, C., Li, X., Fan, Z., Xu, J., Nag, K., Korpeoglu, E., Kumar, S., & Achan, K. (2024). Llm-ensemble: Optimal large language model ensemble method for e-commerce product attribute value extraction. Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval,
Finardi, P., Avila, L., Castaldoni, R., Gengo, P., Larcher, C., Piau, M., Costa, P., & Caridá, V. (2024). The chronicles of rag: The retriever, the chunk and the generator. arXiv preprint arXiv:2401.07883.
Ghani, R., Probst, K., Liu, Y., Krema, M., & Fano, A. (2006). Text mining for product attribute extraction. SIGKDD Explor. Newsl., 8(1), 41–48. https://doi.org/10.1145/1147234.1147241
Grattafiori, A., Dubey, A., Jauhri, A., Pandey, A., Kadian, A., Al-Dahle, A., Letman, A., Mathur, A., Schelten, A., & Vaughan, A. (2024). The llama 3 herd of models. arXiv preprint arXiv:2407.21783.
Grootendorst, M. (2020). KeyBERT: Minimal keyword extraction with BERT. Zenodo. https://doi.org/10.5281/zenodo.4461265
Hu, T., & Zhou, X.-H. (2024). Unveiling llm evaluation focused on metrics: Challenges and solutions. arXiv preprint arXiv:2404.09135.
Jalil, A., & Kobayashi, J. (2022, 27 Nov.-1 Dec. 2022). Efficacy of Local Cache for Performance Improvement of Reliable Data Transmission in Aggregated Robot Processing Architecture. 2022 22nd International Conference on Control, Automation and Systems (ICCAS),
Jiang, A. Q., Sablayrolles, A., Mensch, A., Bamford, C., Chaplot, D. S., Casas, D. d. L., Bressand, F., Lengyel, G., Lample, G., Saulnier, L., Lavaud, L. e. R., Lachaux, M.-A., Stock, P., Scao, T. L., Lavril, T., Wang, T., Lacroix, T., & Sayed, W. E. (2023). Mistral 7B. ArXiv, abs/2310.06825.
Johnny, S., & Jaya Nirmala, S. (2019). Key Phrase Extraction System for Agricultural Documents. In A. B. Gani, P. K. Das, L. Kharb, & D. Chahal, Information, Communication and Computing Technology Singapore.
Khosla, P., Teterwak, P., Wang, C., Sarna, A., Tian, Y., Isola, P., Maschinot, A., Liu, C., & Krishnan, D. (2020). Supervised contrastive learning. Advances in neural information processing systems, 33, 18661-18673.
Kong, A., Zhao, S., Chen, H., Li, Q., Qin, Y., Sun, R., & Bai, X. (2023, July). PromptRank: Unsupervised Keyphrase Extraction Using Prompt. In A. Rogers, J. Boyd-Graber, & N. Okazaki, Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) Toronto, Canada.
Li, X., Sun, L., Ling, M., & Peng, Y. (2023). A survey of graph neural network based recommendation in social networks. Neurocomputing, 549, 126441. https://doi.org/https://doi.org/10.1016/j.neucom.2023.126441
Lin, C.-Y. (2004, July). ROUGE: A Package for Automatic Evaluation of Summaries.Text Summarization Branches Out Barcelona, Spain.
Liu, B., Mazumder, S., Robertson, E., & Grigsby, S. (2023). AI autonomy: Self‐initiated open‐world continual learning and adaptation. AI Magazine, 44(2), 185-199.
Liu, H., Yin, Q., Wang, Z., Zhang, C., Jiang, H., Gao, Y., Li, Z., Li, X., Zhang, C., Yin, B., Wang, W., & Zhu, X. (2023). Knowledge-Selective Pretraining for Attribute Value Extraction. https://doi.org/10.18653/v1/2023.findings-emnlp.542
Mehta, K., Oprea, I., & Rasiwasia, N. (2021). Latex-numeric: Language-agnostic text attribute extraction for e-commerce numeric attributes. arXiv preprint arXiv:2104.09576.
Mihalcea, R., & Tarau, P. (2004, July). TextRank: Bringing Order into Text. In D. Lin & D. Wu, Proceedings of the 2004 Conference on Empirical Methods in Natural Language Processing Barcelona, Spain.
Mikolov, T. (2013). Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781.
Nair, R. P., & Thushara, M. G. (2024). Investigating Natural Language Techniques for Accurate Noun and Verb Extraction. Procedia Computer Science, 235, 2876-2885. https://doi.org/https://doi.org/10.1016/j.procs.2024.04.272
Naveed, H., Khan, A. U., Qiu, S., Saqib, M., Anwar, S., Usman, M., Akhtar, N., Barnes, N., & Mian, A. (2023). A comprehensive overview of large language models. arXiv preprint arXiv:2307.06435.
Ni, J., Abrego, G. H., Constant, N., Ma, J., Hall, K. B., Cer, D., & Yang, Y. (2021). Sentence-t5: Scalable sentence encoders from pre-trained text-to-text models. arXiv preprint arXiv:2108.08877.
OpenAI. (2022). Introducing ChatGPT. OpenAI. Retrieved 7 October 2024 from https://openai.com/index/chatgpt/
Pecar, S., Farkas, M., Simko, M., Lacko, P., & Bielikova, M. (2018, October). NL-FIIT at IEST-2018: Emotion Recognition utilizing Neural Networks and Multi-level Preprocessing. In A. Balahur, S. M. Mohammad, V. Hoste, & R. Klinger, Proceedings of the 9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis Brussels, Belgium.
Polys, N., Mohammed, A., & Sandbrook, B. (2024). Prompt Engineering for X3D Object Creation with LLMs Proceedings of the 29th International ACM Conference on 3D Web Technology, Guimarães, Portugal. https://doi.org/10.1145/3665318.3677159
Prep, A. (2024). 2024 Amazon Seller Statistics: Essential Data You Need to Know. AMZ Prep. Retrieved 16 October 2024 from 2024 Amazon Seller Statistics: Essential Data You Need to Know
Push-Pull. (2022). How Many People Use Amazon in 2022? The Latest Statistics. Push-Pull. Retrieved 16 October 2024 from https://pushpullagency.com/blog/how-many-people-use-amazon/
Putthividhya, D., & Hu, J. (2011). Bootstrapped named entity recognition for product attribute extraction Proceedings of the Conference on Empirical Methods in Natural Language Processing, Edinburgh, United Kingdom.
Qi, C., Zheng, J., Yang, X., Chen, Q., & Wu, M. (2023). Application of deep neural network in the strength prediction of cemented paste backfill based on a global dataset. Construction and Building Materials, 391, 131827. https://doi.org/https://doi.org/10.1016/j.conbuildmat.2023.131827
Radford, A., & Narasimhan, K. (2018). Improving Language Understanding by Generative Pre-Training.
Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2020). Exploring the limits of transfer learning with a unified text-to-text transformer. Journal of machine learning research, 21(140), 1-67.
Raza, S., & Ding, C. (2022). News recommender system: a review of recent progress, challenges, and opportunities. Artificial Intelligence Review, 55(1), 749-800. https://doi.org/10.1007/s10462-021-10043-x
Rezk, M., Alemany, L. A., Nio, L., & Zhang, T. (2019, 8-11 April 2019). Accurate Product Attribute Extraction on the Field. 2019 IEEE 35th International Conference on Data Engineering (ICDE),
Rijn, J. N. v., & Hutter, F. (2018). Hyperparameter Importance Across Datasets Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, London, United Kingdom. https://doi.org/10.1145/3219819.3220058
Sabeh, K., Kacimi, M., Gamper, J., Litschko, R., & Plank, B. (2024). Exploring Large Language Models for Product Attribute Value Identification. arXiv preprint arXiv:2409.12695.
Sazu, M. H. (2022). Does Big Data Drive Innovation In E-Commerce: A Global Perspective? SEISENSE Business Review.
Schulhoff, S., Ilie, M., Balepur, N., Kahadze, K., Liu, A., Si, C., Li, Y., Gupta, A., Han, H., & Schulhoff, S. (2024). The prompt report: A systematic survey of prompting techniques. arXiv preprint arXiv:2406.06608.
Shan–shan, Z., & Zhi–li, W. (2021, 13-15 Aug. 2021). Formal Definition of Pseudo Code and Mapping Rules to Java Code. 2021 IEEE 4th International Conference on Computer and Communication Engineering Technology (CCET),
Shendy, R., & Nalepa, J. (2024). Few-shot satellite image classification for bringing deep learning on board OPS-SAT. Expert Systems with Applications, 251, 123984.
Shinzato, K., Yoshinaga, N., Xia, Y., & Chen, W.-T. (2022, May). Simple and Effective Knowledge-Driven Query Expansion for QA-Based Product Attribute Extraction. In S. Muresan, P. Nakov, & A. Villavicencio, Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers) Dublin, Ireland.
Statista. (2024). eCommerce - Worldwide. Statista. Retrieved 17 August 2024 from https://www.statista.com/outlook/emo/ecommerce/worldwide
Su, Y., Lloyd, L., Chen, X., & Chase, J. G. (2023). Latency mitigation using applied HMMs for mixed reality-enhanced intuitive teleoperation in intelligent robotic welding. The International Journal of Advanced Manufacturing Technology, 126(5), 2233-2248. https://doi.org/10.1007/s00170-023-11198-3
Sun, Y., Qiu, H., Zheng, Y., Wang, Z., & Zhang, C. (2020). SIFRank: A New Baseline for Unsupervised Keyphrase Extraction Based on Pre-Trained Language Model. IEEE Access, 8, 10896-10906. https://doi.org/10.1109/ACCESS.2020.2965087
T., B. (2024). Amazon Statistics: Key Numbers and Fun Facts. AMZSCOUT. Retrieved 16 October 2024 from https://amzscout.net/blog/amazon-statistics/
Touvron, H., Lavril, T., Izacard, G., Martinet, X., Lachaux, M.-A., Lacroix, T., Rozière, B., Goyal, N., Hambro, E., & Azhar, F. (2023). Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971.
Touvron, H., Martin, L., Stone, K., Albert, P., Almahairi, A., Babaei, Y., Bashlykov, N., Batra, S., Bhargava, P., & Bhosale, S. (2023). Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288.
Vaswani, A. (2017). Attention is all you need. Advances in neural information processing systems.
Vinh, N. X., Epps, J., & Bailey, J. (2010). Information Theoretic Measures for Clusterings Comparison: Variants, Properties, Normalization and Correction for Chance. J. Mach. Learn. Res., 11, 2837–2854.
Wang, Q., Yang, L., Kanagal, B., Sanghai, S., Sivakumar, D., Shu, B., Yu, Z., & Elsas, J. (2020). Learning to Extract Attribute Value from Product via Question Answering: A Multi-task Approach Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Virtual Event, CA, USA. https://doi.org/10.1145/3394486.3403047
Wang, Q., Yang, L., Wang, J., Krishnan, J., Dai, B., Wang, S., Xu, Z., Khabsa, M., & Ma, H. (2022, December). SMARTAVE: Structured Multimodal Transformer for Product Attribute Value Extraction. In Y. Goldberg, Z. Kozareva, & Y. Zhang, Findings of the Association for Computational Linguistics: EMNLP 2022 Abu Dhabi, United Arab Emirates.
Wu, J., Chen, X.-Y., Zhang, H., Xiong, L.-D., Lei, H., & Deng, S.-H. (2019). Hyperparameter Optimization for Machine Learning Models Based on Bayesian Optimizationb. Journal of Electronic Science and Technology, 17(1), 26-40. https://doi.org/https://doi.org/10.11989/JEST.1674-862X.80904120
Xu, H., Wang, W., Mao, X., Jiang, X., & Lan, M. (2019). Scaling up Open Tagging from Tens to Thousands: Comprehension Empowered Attribute Value Extraction from Product Title. Annual Meeting of the Association for Computational Linguistics,
Xu, L., Zhang, C., Li, X., Shang, J., & Choi, J. D. (2023). Towards open-world product attribute mining: A lightly-supervised approach. arXiv preprint arXiv:2305.18350.
Yang, L., Wang, Q., Wang, J., Quan, X., Feng, F., Chen, Y., Khabsa, M., Wang, S., Xu, Z., & Liu, D. (2023, July). MixPAVE: Mix-Prompt Tuning for Few-shot Product Attribute Value Extraction. In A. Rogers, J. Boyd-Graber, & N. Okazaki, Findings of the Association for Computational Linguistics: ACL 2023 Toronto, Canada.
Yang, L., Wang, Q., Yu, Z., Kulkarni, A., Sanghai, S., Shu, B., Elsas, J., & Kanagal, B. (2022). Mave: A product dataset for multi-source attribute value extraction. Proceedings of the fifteenth ACM international conference on web search and data mining,
Zahedi, L., Mohammadi, F. G., Rezapour, S., Ohland, M. W., & Amini, M. H. (2021). Search algorithms for automated hyper-parameter tuning. arXiv preprint arXiv:2104.14677.
Zhang, L., Chen, Q., Wang, W., Deng, C., Zhang, S., Li, B., Wang, W., & Cao, X. (2022, May). MDERank: A Masked Document Embedding Rank Approach for Unsupervised Keyphrase Extraction. In S. Muresan, P. Nakov, & A. Villavicencio, Findings of the Association for Computational Linguistics: ACL 2022 Dublin, Ireland.
Zhang, S., Yao, L., Sun, A., & Tay, Y. (2019). Deep Learning Based Recommender System: A Survey and New Perspectives. ACM Comput. Surv., 52(1), Article 5. https://doi.org/10.1145/3285029
Zhang, T., Kishore, V., Wu, F., Weinberger, K. Q., & Artzi, Y. (2019). Bertscore: Evaluating text generation with bert. arXiv preprint arXiv:1904.09675.
Zhang, X., Zhang, C., Li, X., Dong, X. L., Shang, J., Faloutsos, C., & Han, J. (2022). OA-Mine: Open-World Attribute Mining for E-Commerce Products with Weak Supervision Proceedings of the ACM Web Conference 2022, Virtual Event, Lyon, France. https://doi.org/10.1145/3485447.3512035
Zhao, W. X., Zhou, K., Li, J., Tang, T., Wang, X., Hou, Y., Min, Y., Zhang, B., Zhang, J., & Dong, Z. (2023). A survey of large language models. arXiv preprint arXiv:2303.18223.
Zheng, G., Mukherjee, S., Dong, X. L., & Li, F. (2018). Opentag: Open attribute value extraction from product profiles. Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining,
Zhu, T., Wang, Y., Li, H., Wu, Y., He, X., & Zhou, B. (2020). Multimodal Joint Attribute Prediction and Value Extraction for E-commerce Product. Conference on Empirical Methods in Natural Language Processing,