My我的 review on MSN Search檢討對MSN搜索 elicited the following response from引起下列反應 MSN Search’s Weblog MSN搜索的博客 : ：

Angsuman Chakraborty日由Angsuman Chakraborty thinks that we released the Beta too early.認為我們公佈的測試版還為時過早。 He also questions the veracity of our site descriptions.他還問題的真實性，我們的網站上說明。 I want to respond to this by re-iterating that there is no human intervention in our Index; all results are based on our algorithms.我想就此作出回應，重新迭代是有沒有人為干預，在我們的索引;所有結果都是基於我們的算法。

In response I posted to the same weblog:在回應我張貼到相同的博客：

# re: One week out and here’sa roundup ＃回复：一周及以下綜述
Angsuman Chakraborty日由Angsuman Chakraborty
> I want to respond to this by re-iterating that there is no human intervention in our Index; all results are based on our algorithms. >我想就此作出回應，重新迭代是有沒有人為干預，在我們的索引;所有結果都是基於我們的算法。

I am happy to know that.我很高興知道這一點。
What bothers me is that certain sites seems to have rather strange description, something which is definitely not derived from its contents or meta tags.是什麼困擾我的是，某些網站上似乎有很奇怪的描述，這絕對不是來自其內容或中繼標記。 An example which I quoted there for Oracle -一個例子，我所引述的有甲骨文-
“Silicon Valley software giant supplies applications for enterprise information management. “矽谷軟件巨頭用品的申請，企業信息化管理。 Features demos and upgrade downloads.”功能演示和升級下載“ 。

It is however not so with many other sites.這是並非如此，但與其他許多網站。 I can clearly see that it normally places a great emphasis on meta tags, specially the description tag to provide summary information.我可以清楚地看出，它通常是地方十分重視META標記，特別是描述標記，以提供簡要資料。

As you said it is based on algorithms.正如你所說的它是基於算法。 I don’t disagree with that.我不同意這種說法。 However I wonder if the algorithms take as its input some text file (or other information source) which contains specific instructions & descriptions regarding some selected sites不過，我不知道如果算法作為它的輸入一些文本文件（或其他信息源） ，其中包含了具體的指示與說明，對於一些選定地點 ： ）

The reason for this assumption is because certain site descriptions are clearly not content based.這種情況的原因的假設，是因為在某個網站上說明，顯然不是基於內容的。

Looking forward to your thoughts on this.期待著您的想法就這。

Sometime has passed since the review, so I decided to take a look again at MSN Search.一段時間以來，已通過審查，所以我決定採取再看看MSN搜索。

I looked back again at the old query “Sun” and here’s what I got MSN search.我翻查回再次在舊查詢“太陽”和這裡的什麼我得到的MSN搜索。

Sun Microsystems Sun Microsystems公司

Developer of the Java language offers information on new products, and update information on existing ones.開發的Java語言提供的資料，新產品，並更新資料，現有定居點。

However on然而， www.sun.com the description is:描述是：

Sun Microsystems, Inc. The Network Is The Computer[tm].為Sun Microsystems ， Inc網絡就是計算機[技術備忘錄] 。

Nowhere on the site is the verbiage (or even anything remotely near to it) that is used in your query result summary.沒有在網站上是多系（或什至任何遠程附近的話）是用來在您的查詢結果的概要。

I did a google search and couldn’t find anywhere similar description about Sun.我做了Google搜尋，並無法找到任何類似的描述關於孫。 It is a common knowledge that Sun is a hardware company.它是一種常見的知識，孫是一家硬件公司。 It is the creator of Java.這是造物主的Java 。 However that is not how you would primarily define Sun!不過，這不是你如何將主要界定孫！

Interestingly if you search about有趣如果您搜索約 my company我的公司 or many other sites, you will get a description taken from the meta tag description .或許多其他網站，你將獲得一個描述採取從中繼標記說明 。 It is obvious that the meta tag is given undue importance by MSN Search Beta, unlike Google.這是很明顯的中繼標記是過分的重要性， MSN搜索測試版，與Google不同。 However this rule is flouted for certain sites like Sun, Oracle etc.但是這條規則是藐視某些網站，像太陽，甲骨文等。

Am I wrong in assuming that MSN Search query summary for certain sites are written by humans (and often with incorrect understanding of the industry, or is it intentional?), which is then picked by search engine algorithm in preference to automatically generated summary?我在錯誤的假設MSN搜索查詢的概要某些網站的書面人類（往往與不正確了解業界，抑或是故意？ ） ，然後挑選出來的搜索引擎算法在優先自動生成的摘要嗎？

You decide.由你決定。 Let me know your thoughts.讓我知道您的想法。