相关性成就个性化搜索新时代

    在相关性搜索的应用上，旅行和在线零售行业走在最前列，不过其它行业也不甘落后。比如谷歌（Google）和必应（Bing）等搜索引擎也开始采用这一理念，开始根据用户的搜索历史和个人喜好，向不同人群显示不同的搜索结果。另外近年来非常热门的团购业也在努力整合相关性搜索，因为消费者越来越需要一个能够提供折扣的购物平台（而不是向吃素的人提供牛排餐厅的优惠券）。围绕着用户意图能够构建起来的背景信息越多，显示的搜索结果就会越精确。

    不过到目前为止，还是旅行业最能显示相关性搜索的能力和价值。

    今天，人们在打算出行的时候，往往要登陆好几十个网站，搜索符合要求的航班和酒店。每次都要颇费一番周折——尤其是每年的这个时候，机票和酒店更是贵得吓人，以至于消费者无论花多大的力气，也要货比N家，找出其中最便宜的。

    摆在大多数旅行者面前的一个问题是：一般说来，如果在网络上搜索一个航班的信息，会跑出来1,000多个难以区别的、混乱无序的搜索结果。现在许多商业旅行平台，甚至是有些消费者网站都开始使用语义数据、统计建模和机器学习等手段来解决这一问题。有些高级平台可以迅速扫描用户的相关数据——比如可能的航班时间、可接受的价格区间、用户需要的机内设施，以及用户喜欢的航空公司等——然后为每名用户重点展示最佳选择。

    这些工具并不会把成千上万个其它搜索结果排除掉，而是会把所有搜索结果组织起来，优先显示相关性最强的选项。这种组织方法意味着用户可以轻易地对背景环境进行修改，比如今天可以想选择“廉价旅行”，明天又变成“不计成本”。但不论采取哪种情况，相关性搜索引擎都能快速地分析人们的个人偏好，提供最好的选择。

    消费者对相关性搜索能力的需求已经存在了一段时间了。各大搜索引擎和企业早就在尝试如何利用消费者对个性化信息的需求，现在相关性搜索终于开始走向成熟。在许多方面，相关性搜索都是对消费者行为变化方式的一种回应。以往消费者在购物或订票时会先打开黄页，但这已经是老黄历了。数字时代的消费者生活在网络上，他们不能容忍收件箱里的垃圾广告邮件，也不愿在吃饭时冷不防接到促销电话。他们希望能够控制自己收到的信息，以及何时收到这些信息。未来消费者在相关性和个性化问题上还会变得更加苛刻。

    Travel and online retailers are ahead of the curve, but others are embracing relevance-powered search. Search engines like Google and Bing are also beginning to adopt the concept -- serving up different results for different people, based on search history and personal preferences. And fast-growing daily deal businesses are working hard to incorporate relevance as consumers increasingly demand platforms that provide appropriate discounts (instead of offering steakhouse coupons to vegetarians). The more contexts that can be built around a user's intent, the more refined the result set can, and will, be.

    But travel is by far one of the best ways to showcase the power and value of relevance.

    Today, travelers spend hours searching dozens of websites for flights and hotels that meet their requirements. It's a major hassle every time they need to travel – especially this time of year, when air fare is so high that consumers will do everything in their power to get the best deal possible.

    The problem faced by most travelers is that a typical flight search will yield 1,000+ nearly indistinguishable and unorganized results. Many business travel platforms – and even some consumer websites – are starting to solve this problem through the use of semantic data, statistical modeling and machine learning. Advanced platforms can quickly scan available data on a user – like the times they can fly, their price range, the in-flight amenities they desire and their preferred airlines – and highlight the best options for each user.

    These tools won't eliminate the tens of thousands of results available to the user, but they will organize and structure them so the most relevant options are presented first. This structuring means that the context of a search can easily be modified whether the user is in a 'budget trip mood' today, or a 'no expense spared mood' tomorrow. Either way, relevance engines can quickly analyze your personal preferences to serve up the best options.

    The desire for relevance-powered search capabilities has been around for some time. Companies and search engines have long been trying to take advantage of the buyer's need for personalized information, and it's finally coming to fruition. In a lot of ways, it's a reaction to how consumer behavior is changing. While the old-school buyer might start the process in the yellow pages, today's digital-aged buyer lives online. They're far less tolerant of unwanted advertisements spamming their inboxes and cold calls during dinner time – and want control over the information they receive, and how they receive it. Tomorrow's buyer will be even more demanding when it comes to relevance and personalization.