原本想做精緻些，整理出一週閱讀素材的處理方法，每日一篇，留日後反思。現在就跟著小編一起來了解一下《時尚玄學或遭資料瓦解》吧。
In the film “The Devil Wears Prada”, the character of Miranda Priestly, whose role is based on a feared Vogue editor, scolds her new assistant for not understanding fashion. Fashion, she tells her, is whatever a select group of designers and critics says it is. What she does not say, however, is that their judgments are themselves often influenced by another group: fashion forecasters, who predict what will be “in”. Might these seers of style in turn be undone by artificial intelligence (AI)?
電影《穿普拉達的女王》中的米蘭達取材於一位令人聞風喪膽的Vogue主編，影片中她訓斥新來的助理不懂什麼是時尚，並告訴助理，一群頂尖設計師和評論家湊到一起，說時尚是什麼，時尚就是什麼。但她未提及，這些人的判斷還常常受到另一群人的影響:預言未來流行趨勢的時尚預測師。那會不會有一天也輪到這些潮流先知們被人工智慧(AI)取代呢?
Fashion forecasting has always been a peculiar profession. The business came into its own in Paris in the 1960s when agencies began releasing “trend books”, collections of fabrics and design ideas. Retailers use these books for inspiration as they put together designs.
時尚預測向來是個玄乎的行當。上世紀六十年代在巴黎，這一行自立門戶，開始發行“流行趨勢解讀”，內容整合了各種流行面料和設計亮點，零售商在拼湊設計時讀這些書以尋求靈感。
The biggest of these forecasting firms is WGSN, with a market share of 50%. It employs 150 forecasters who scour the world’s catwalks, bars and clubs to spot the next big thing. Their findings are then combined with other data, from economic indicators to political sentiment. Petah Marian, a senior editor at WGSN, is confident that the methodology works. She says her colleagues often exclaim “I forecast that!” when visiting clothing shops.
最大的時尚預測公司是WGSN，佔有50%的市場份額。其麾下有150名時尚預測師，在全世界的秀場，酒吧，夜店裡沙裡淘金，尋找下一個時尚熱點。繼而他們的發現將與其他資料整合起來，從經濟指標到政治情感，無一例外。WGSN高階編輯佩塔·瑪麗安對這套方法信心滿滿。她說，她的同事們逛服裝店的時候經常會興奮地叫“我預測到了這個!”
Ms Marian’s confidence may seem surprising, given the lack of clear correlations between fashion and macroeconomic data. Not much evidence supports the theory of George Taylor, an economist, that hemlines rise with stocks, and Leonard Lauder’s suggestion that lipstick sales increase during a downturn. Even the cofounder of WGSN, Marc Worth, who sold the firm to set up a rival service, once stated: “Nobody can really predict or forecast trends.” If forecasters can claim accuracy rates of up to 80%, it is because their predictions are often self-fulfilling. Most major retailers buy trend books. For designers, they are a form of insurance: as long as they are widely used, the risk of being wildly out of step with the market is modest.
瑪麗安女士的信心看上去有點不可思議，畢竟時尚和巨集觀經濟資料之間並無明顯的聯絡。經濟學家喬治·泰勒的理論稱裙子越短，股市越好，這一說法也沒什麼證據支援，萊納德·勞德所說的經濟下滑時口紅銷量會上漲亦是如此。WSGN創始人之一的馬克·沃斯，賣掉WSGN後成立了一家競爭機構，他曾說：“沒有人能真正預測流行趨勢。”如果預測師能確保準確率在80%以上，則是因為他們的預測通常可以自我實現。主要零售商是流行趨勢預言的'受眾。對設計師們來說，這些預言提供了某種形式的保障：只要預言廣而泛之，自己過分偏離市場的機率就微乎其微。
The business of forecasting is menaced by data-driven analysis, however. The clothing industry’s supply chain is becoming more digital and more flexible: Inditex and H&M, for example, aim to take an idea and turn it into a finished product ready for mass production in two weeks. In response, forecasting agencies are making use of data collated from retailers’ IT systems and have added short-term predictions to their portfolio of services. In 2013 WGSN launched INstock, a retail-analytics service, which uses past sales figures to predict upcoming bestsellers. EDITED, a competing service, provides “solid metrics” in fashion, claiming to use machine learning, an AI technique, in order to predict short-term sales trends.
然而，以資料為主的分析法對商業預測構成了威脅。在時裝行業，供應鏈越發數字化，體系也更加靈活。像印地迪克和H&M兩大經銷商就打算一旦確定某個想法，就得在兩週之內製出成品，用於大規模生產。相應地，預測機構對零售商資訊系統中的資料整理利用，並由此在其服務專案中添加了短期預測這一項。2013年，WGSN釋出了INstock軟體，該軟體運用以往的銷售資料對未來暢銷品進行預測，因此可用於零售分析。INstock的競爭對手EDITED號稱運用人工智慧中的機器學習法，能對短期銷售走勢進行預測，從而為時尚業提供“可靠準則”。
Such offerings notwithstanding, the marriage of AI and fashion is still in its infancy. A study in 2014 found that the best predictive models get it wrong nearly half the time. But forecasters are likely to face rising competition as technology firms enter the market. Google, an online giant, now has a “Trend spotting” division. It releases a regular “Fashion Trends Report” based on the firm’s vast trove of search data. So far the results are basic: in 2016 slim “mom jeans” were on the rise while baggier “boyfriend jeans” were on the way out. But Olivier Zimmer, the project’s data scientist, says that the goal is to produce more sophisticated combinations of search and other data.
人工智慧和時尚的結合，雖說孕育了上述產品服務，卻仍處於初級階段。2014年的一項研究發現，即便最好的預測模型，出錯率仍接近百分之五十。隨著技術公司的介入，預測機構還可能面臨更大的挑戰。網路巨頭“谷歌”現在建立了“潮流觀測”部。該部門基於谷歌強大的搜尋資料庫，定期釋出“時尚潮流報告”。目前來講，報告得出的結論尚屬淺顯，比方說裡面會給出“2016年，緊身的“媽媽牛仔褲”興起而寬鬆的“男友牛仔褲”正逐漸落伍”這樣的結論。但負責這些報告的資料學家奧利維·季默稱，更加科學合理地結合搜尋資料與其他資料才是目的所在。
【本文作者：徐州七中彭向梅。(公眾號：草根英語行思教)】
本文為原創文章，版權歸作者所有，未經授權不得轉載!
