CN101080709A - Variable personalization of search results in a search engine - Google Patents

Abstract

A search engine provides personalized ranking of search results. A user interest profile identifies topics of interest to a user. Each topic is associated with one or more sites and a boost value that can be used to increase the information retrieval score of any document from that site. Providing search results from any search to the user, wherein the ranking of the results is variably controlled. The results may be ranked by their non-boosted information retrieval score, so as not to reflect personalization, or by their fully or partially boosted information retrieval score. This enables the user to selectively control the extent to which their interests affect the ranking of the documents.

Description

Translated fromChinese

在搜索引擎中对搜索结果的可变个性化Variable personalization of search results in search engines

相关申请案交叉参照Related Application Cross Reference

本申请案根据35U.S.C.§119(e)主张基于2004年3月29日提出申请的第60/557,783号申请案及2004年12月1日提出申请的第11/002,474号申请案的权利，上述申请案均以引用方式并入本文中。本申请案还以引用方式将第10/646,331号(2003年8月22日提出申请)及第10/641,489号(2003年8月14日提出申请)美国专利申请案并入本文中；所有上述申请案均与本申请案共同拥有。This application asserts rights under 35 U.S.C. § 119(e) based on Application Nos. 60/557,783, filed March 29, 2004, and 11/002,474, filed December 1, 2004, All of the above applications are incorporated herein by reference. This application also incorporates by reference U.S. Patent Application Nos. 10/646,331 (filed August 22, 2003) and 10/641,489 (filed August 14, 2003); all of the above applications are jointly owned with this application.

技术领域technical field

本申请案大体而言涉及在线搜索引擎，且具体而言涉及对搜索结果的个性化。This application relates generally to online search engines, and in particular to personalizing search results.

背景技术Background technique

在线搜索引擎技术的当前技术水平在其检索可回应于查询项的文档(例如网页或其某些部分、图像等等)的能力方面非常先进。现今的搜索引擎可快速检索与查询项相匹配的特定文档。然而，当前的搜索引擎常常返回虽然准确地对应于特定查询项但根本反映不出用户的基本兴趣的文档。因此，两个不同用户(其中一个对体育非常感兴趣，而另一个对政治感兴趣)在一给定查询中将得到完全相同的结果，例如“棒球中的药检”，即使第一个用户可能对了解哪些队已实施了药检更感兴趣，而后一个用户可能对了解与在体育中进行药检相关的立法更感兴趣。因此，需要提供一种根据用户的兴趣使搜索结果个性化的机理及方法。The current state of the art of online search engine technology is very advanced in its ability to retrieve documents (eg, web pages or portions thereof, images, etc.) that respond to query terms. Today's search engines quickly retrieve specific documents that match query terms. However, current search engines often return documents that, while accurately corresponding to specific query terms, do not reflect the user's underlying interests at all. Thus, two different users (one of whom is very interested in sports and the other interested in politics) will get exactly the same results for a given query, such as "drug test in baseball", even though the first user may While more interested in knowing which teams have implemented drug testing, the latter user may be more interested in knowing legislation related to drug testing in sports. Therefore, it is necessary to provide a mechanism and method for personalizing search results according to user interests.

此外，虽然用户可能具有可适用于处理其搜索查询的特定的一组兴趣，但用户可能并不想总是使这些兴趣影响查询结果。因此，将希望提供一种使用户可对他的兴趣影响一给定查询的结果的程度进行可变调整的机理及方法。Furthermore, while a user may have a specific set of interests that may be applicable to addressing their search query, the user may not always want these interests to influence the query results. Accordingly, it would be desirable to provide a mechanism and method by which a user can variably adjust the degree to which his interests affect the results of a given query.

发明内容Contents of the invention

本发明提供使一用户能够对他或她的个人兴趣影响一给定查询的结果表示形式的程度进行可变调整的系统及方式。本发明具有多个实施例，包括一种用于提供一搜索引擎的系统、一种用于处理一搜索查询的方法及一种供用户对所述影响程度进行可变调整的用户界面。The present invention provides systems and methods that enable a user to variably adjust the degree to which his or her personal interests affect the result presentation for a given query. The invention has multiple embodiments, including a system for providing a search engine, a method for processing a search query, and a user interface for variably adjusting the degree of influence by a user.

从用户的观点而言，本发明提供一种用于使搜索结果个性化的机理。所述用户可规定一组兴趣(例如所述用户感兴趣的一组主题或类别)作为一兴趣配置文件。然后，所述用户将一搜索查询提交至一搜索引擎。所述搜索引擎对与所述查询相关的文档进行检索并提供这些文档作为搜索结果。首先(虽然未必尽然)，在不考虑用户兴趣的情况下对所述搜索结果进行排列。此使用户能够看到所述文档的相关性而不受他的个人兴趣影响。用户可按可变方式控制对搜索结果集合中文档的排列。在一较佳实施例中，此通过直接操纵图形控制元件(一种个性化控制)来进行。此图形元件可具有不同的图形表示形式，且在一实施例中将其表示成一滑标。调整所述控制元件便能根据他的兴趣配置文件中所表达的用户兴趣来调整所述结果的排列程度。响应于所述控制元件的位置变化，对所述搜索结果进行重新排列，从而使所述用户能够看到他的兴趣如何影响对结果的排序。From the user's point of view, the present invention provides a mechanism for personalizing search results. The user may specify a set of interests (eg, a set of topics or categories of interest to the user) as an interest profile. The user then submits a search query to a search engine. The search engine retrieves documents relevant to the query and provides these documents as search results. First (although not necessarily), the search results are ranked without regard to user interests. This enables the user to see the relevance of said documents regardless of his personal interests. The user may control the arrangement of documents in the search result set in a variable manner. In a preferred embodiment, this is done by direct manipulation of graphical control elements (a type of personalization control). This graphical element can have different graphical representations, and in one embodiment it is represented as a slider. Adjusting the control element adjusts the ranking of the results according to the user's interests expressed in his interest profile. In response to a change in the position of the control element, the search results are rearranged, thereby enabling the user to see how his interests affect the ordering of the results.

在一实施例中，搜索结果的排列会起增加与主题目录相关联的文档的相关性得分的作用。所述目录中的每一主题均与一组被认为与所述主题相关的文档(例如网站)相关联。对于每一文档(同样为网站)，均提供一指示所述站点与所述主题的相关程度的提升值。一给定站点内的所有页面均获得该站点的提升值。所述提升值可理解为一表示所述站点与所述主题的相关性的比例因数。在查询处理期间，返回该组文档并进行计分而不考虑其包含于所述主题目录中。另外，确定每一文档的单独得分，其中所述文档的基本相关性得分是通过其相关联的提升值(如果有)来进行调整，例如通过将所述相关性得分乘以所述提升值，从而按比例缩放所述文档与所述查询的相关性。一般而言，只有属于所述目录中所列站点的那些文档才会具有一提升的相关性得分，而所述搜索结果中的其它文档将不会具有所述提升的相关性得分。因此，根据文档的提升的相关性得分对文档进行重新排列会随个性化控制的位置而变化-所述个性化控制的位置从毫无影响(其中所述提升值对所述相关性得分毫无影响)到完全影响(其中将每一文档的提升值100％地应用至所述相关性得分)、及任意数量的中间应用级。因此，用户可通过调整所述控制的位置来立即对所述搜索结果中的文档进行重新排列。In one embodiment, the ranking of the search results acts to increase the relevance score of the documents associated with the subject category. Each topic in the directory is associated with a set of documents (eg, web sites) that are considered to be related to the topic. For each document (also a website), a boost value is provided indicating how relevant the site is to the topic. All pages within a given site receive that site's boost value. The boost value can be understood as a scaling factor representing the relevance of the site to the topic. During query processing, the set of documents is returned and scored regardless of their inclusion in the subject catalog. Additionally, an individual score is determined for each document, wherein the document's base relevance score is adjusted by its associated boost value (if any), for example by multiplying the relevance score by the boost value, The relevance of the document to the query is thereby scaled. In general, only those documents belonging to sites listed in the directory will have a boosted relevance score, while other documents in the search results will not have the boosted relevance score. Thus, rearranging documents according to their boosted relevance scores varies with the position of the personalization control - the position of the personalization control never having any effect (wherein the boost value has no effect on the relevance score Impact) to full impact (where 100% of each document's boost value is applied to the relevance score), and any number of intermediate levels of application. Thus, the user can instantly rearrange the documents in the search results by adjusting the position of the control.

一系统实施例包括一用于向用户提供主题(类别)的主题目录、一用于响应于用户的查询而搜索带索引的网站的搜索引擎及索引、及一用于根据用户的兴趣为所述搜索结果确定所述搜索结果的至少一种(及较佳一组)排列的个性化组件。一用户界面实施例包括一具有各种主题的主题目录、及其中所述用户可选择兴趣主题的若干图形控制器(例如复选框)、及一允许用户根据他的兴趣配置文件来对所述搜索结果排列进行调整的图形控制器。One embodiment of the system includes a subject directory for providing topics (categories) to users, a search engine and index for searching indexed websites in response to user queries, and a system for providing information to the users based on their interests. The search results determine at least one (and preferably one set) permutation personalization component of said search results. One user interface embodiment includes a topic directory with various topics, and several graphical controls (such as checkboxes) where the user can select topics of interest, and a user interface that allows the user to select the topics according to his interest profile. Search results arrangement adjusted for graphics controllers.

接下来，参照各图式及图解以及技术信息来阐述本发明。所述图式只是出于举例说明的目的来描绘本发明各实施例。所属领域的技术人员通过下文说明将易知，也可采用所示及所述结构、方法及功能的替代实施例，此并不背离本发明的原理。Next, the present invention is explained with reference to the drawings and diagrams and technical information. The drawings depict various embodiments of the invention for purposes of illustration only. Those skilled in the art will readily appreciate from the following description that alternative embodiments of the structures, methods, and functions shown and described may be employed without departing from the principles of the invention.

附图说明Description of drawings

图1为一用于一搜索引擎的入口网页。FIG. 1 is a portal page for a search engine.

图2为用于编辑一用户的个人主题配置文件的网页。Figure 2 is a web page for editing a user's personal theme profile.

图3为另一主题及副主题网页。Figure 3 is another topic and subtopic webpage.

图4为另一副主题网页。Figure 4 is another sub-theme webpage.

图5显示带有查询项“stanford”的搜索引擎的入口网页。FIG. 5 shows the entry web page of the search engine with the query term "stanford".

图6显示关于查询项“stanford”的初始搜索结果以及一用于在对所述搜索结果进行排列时控制所述用户的兴趣配置文件的影响程度的控制器608。FIG. 6 shows initial search results for the query term "stanford" and acontroller 608 for controlling the degree of influence of the user's interest profile in ranking the search results.

图7显示在所述用户已将控制器608调整至一中间位置后图6所示搜索结果的排列。FIG. 7 shows the arrangement of the search results shown in FIG. 6 after the user has adjusted thecontroller 608 to an intermediate position.

图8显示在所述用户已将控制器608调整至一最大位置后图6所示搜索结果的排列。FIG. 8 shows the arrangement of the search results shown in FIG. 6 after the user has adjusted thecontroller 608 to a maximum position.

图9显示在所述用户删除一“Computer”主题并选择一“Music”主题后的一主题网页。Figure 9 shows a topic web page after the user deletes a "Computer" topic and selects a "Music" topic.

图10显示在图9中改变主题兴趣后图6所示搜索结果的排列。FIG. 10 shows the arrangement of the search results shown in FIG. 6 after changing the subject interests in FIG. 9 .

具体实施方式Detailed ways

现在参见图1，图中显示一搜索引擎的入口(或登陆页面)100。所述搜索引擎包括一其中用户输入一搜索查询项的查询输入栏102。一搜索按钮106将所述查询提交给所述搜索引擎进行处理。所述用户在首次访问所述搜索引擎时，将还没有对一用于使搜索结果个性化的用户配置文件进行界定，且对此显示一消息。所述用户可访问编辑Edit Profile(配置文件)链接104以便创建其个人兴趣。Referring now to FIG. 1, an entry (or landing page) 100 for a search engine is shown. The search engine includes aquery input field 102 in which a user enters a search query. A search button 106 submits the query to the search engine for processing. The user, when first accessing the search engine, will not have defined a user profile for personalizing search results, and a message is displayed to this effect. The user can access the Edit Profile (configuration file) link 104 in order to create their personal interests.

点击编辑链接104便能访问图2中所示页面200。此页面显示一主题目录202的顶层主题204。所述目录中的主题可为(例如)在存在于http://dmoz.org/处的OpenDirectory Project(开放式目录计划)中所规定的主题或者可为某一其它主题目录。(在本揭示内容中，用语“主题”与“类别”可互换，“副主题”与“子类别”也可互换。)所述顶层主题链接可具有一组由箭头206指示的副主题。在任一时刻，用户均可通过点击StartSearching(开始搜索)按钮208来开始一搜索。点击一特定顶层主题便会将用户带到一与所述主题的副主题相关联的页面。Clicking on the edit link 104 accesses thepage 200 shown in FIG. 2 . This page displays a top-level theme 204 of atheme directory 202 . The topics in the directory may be, for example, those specified in the OpenDirectory Project, which exists athttp://dmoz.org/ , or may be some other directory of topics. (In this disclosure, the terms "topic" and "category" are interchangeable, as are "subtopic" and "subcategory.") The top-level topic link may have a set of subtopics indicated byarrows 206 . At any time, the user can start a search by clicking on theStartSearching button 208 . Clicking on a particular top-level topic will take the user to a page associated with that topic's subtopics.

图3图解说明其中用户已点击页面200上的主题“Computer”的此步骤的结果。如图3中所示，主题“Computer”的副主题302已显示在右手的列中。紧挨着每一副主题302的是一复选框314。用户可选择复选框314来指示他对一特定副主题感兴趣。或者，用户可针对整个副主题本身“Computer(General)”选择复选框316。此指示用户对主题“Computer”的所有副主题感兴趣。用户所选择的该组主题可称作用户配置文件、个性化配置文件或兴趣配置文件。在图3的左侧，在用户选择了一个或多个兴趣主题后，在兴趣栏306中列出这些主题。紧挨着每一主题名称的是一使用户能够分别删除各个主题的删除链接308。用户也可通过选择“Delete all preferences”链接310来删除所有主题。FIG. 3 illustrates the result of this step where the user has clicked on the topic "Computer" onpage 200 . As shown in FIG. 3, asubtopic 302 of the topic "Computer" has been displayed in the right-hand column. Next to each subtopic 302 is acheck box 314 . A user may select acheckbox 314 to indicate that he is interested in a particular subtopic. Alternatively, the user may select checkbox 316 for the entire subtopic itself, "Computer (General)". This indicates that the user is interested in all subtopics of the topic "Computer". The set of topics selected by the user may be referred to as a user profile, a personalization profile, or an interest profile. On the left side of FIG. 3 , after the user has selected one or more topics of interest, the topics are listed in aninterest column 306 . Next to each topic name is adelete link 308 that enables the user to delete each topic individually. The user may also delete all topics by selecting the "Delete all preferences"link 310.

所述副主题本身可具有任一数量的附加副主题层。例如，图4图解说明副主题“Games”进一步具有副主题402“Board Games”、“Card Games”等等。这些游戏响应于用户点击紧挨着“Games”主题标签的副主题箭头206而呈现给用户。同样，用户可通过选择复选框404或各个副主题复选框406中的任何一个复选框来选择整个子类别。在一实施例中，除顶层主题204外，还有两个副主题层。此会提供足够的主题覆盖宽度及间隔尺寸，而不会施加不必要的计算复杂度。The subtopic itself may have any number of additional subtopic layers. For example, FIG. 4 illustrates that the subtopic "Games" further has subtopics 402 "Board Games," "Card Games," and so on. These games are presented to the user in response to the user clicking on thesubtopic arrow 206 next to the "Games" hashtag. Likewise, a user may select an entire subcategory by selecting checkbox 404 or any of the individual subtopic checkboxes 406 . In one embodiment, in addition to the top-level theme 204, there are two levels of sub-themes. This provides sufficient topic coverage width and spacing size without imposing unnecessary computational complexity.

重新参见图3，用户已指示他通常对计算机感兴趣。通过点击Start Searching(开始搜索)按钮208会再次调用搜索页面100，搜索页面100不再显示请求用户创建个性化配置文件的消息。这时，用户在所述搜索栏中输入一搜索查询，此处所述查询为图5中所示的“Stanford”。搜索引擎对其索引实施搜索并提供搜索结果。图6图解说明初始搜索结果页面600。此外，搜索结果602包括一组分别由一URL 617指示的文档604连同一通往所述页面的链接615及一向用户显示所述文档上下文的一部分的页面文本片断619。靠近所述页面顶部的是个性化控制器608。在本申请案的上下文中，“文档”应理解为可由一搜索引擎加索引并检索的任一类型的媒体，包括网页、图像、多媒体文件、文本文档、PDF等等。使用通用术语“文档”并不意含着限制本发明的范围。Referring back to Figure 3, the user has indicated that he is generally interested in computers. By clicking the Start Searching (start searching)button 208, the search page 100 will be invoked again, and the search page 100 will no longer display the message requesting the user to create a personalized configuration file. At this time, the user inputs a search query in the search bar, where the query is "Stanford" as shown in FIG. 5 . A search engine conducts searches on its index and provides search results. FIG. 6 illustrates an initial search results page 600 . Additionally, search results 602 include a set ofdocuments 604 each indicated by aURL 617 along with alink 615 to that page and a snippet ofpage text 619 that displays a portion of the document's context to the user. Near the top of the page is apersonalization control 608 . In the context of this application, "document" should be understood as any type of media that can be indexed and retrieved by a search engine, including web pages, images, multimedia files, text documents, PDFs, and the like. Use of the generic term "document" is not intended to limit the scope of the invention.

个性化控制器608使用户能够根据他的兴趣(较佳如所述个性化配置文件中所表示)来对搜索结果的各个文档/页面的排列加以调整。所述个性化控制器可解释为控制用户的兴趣配置文件对搜索结果的排列的影响或作用的程度。图6图解说明在一其中控制器608为一滑标型控制器的实施例中的控制器608，控制器608包括一在最小值610至最大值614之间沿线性维度移动的控制手柄610。此类型的控制器608通常非常直观且易于用户了解其操作方式。替代实施例包括同样在最小与最大值之间提供可变调整的刻度盘类型控制器、及若干分别具有一相关的预定调整量的无线电按钮(例如五个分别与0％、20％、40％、80％及100％相关联的无线电按钮)。用户界面设计领域的技术人员将了解，也可采用其它类型的控制器来达到基本相同的效果。Thepersonalization controller 608 enables the user to adjust the arrangement of the individual documents/pages of the search results according to his interests (preferably as expressed in the personalization profile). The personalization controller may be interpreted as controlling the degree to which a user's interest profile influences or contributes to the ranking of search results. FIG. 6 illustrates thecontroller 608 in an embodiment in which thecontroller 608 is a slider-type controller that includes acontrol handle 610 that moves along a linear dimension between aminimum value 610 and amaximum value 614 . This type ofcontroller 608 is usually very intuitive and easy for the user to understand how it operates. Alternative embodiments include a dial-type control that also provides variable adjustment between minimum and maximum, and several radio buttons each with an associated predetermined amount of adjustment (e.g., five corresponding to 0%, 20%, 40%, respectively). , 80%, and 100% associated radio buttons). Those skilled in the art of user interface design will appreciate that other types of controllers can be used to achieve substantially the same effect.

在图6所示的实例中，个性化控制器608一直设定至其最小值612。因此，根据搜索结果与查询的相关性及与查询无关的信息(例如PageRank)来对相应地搜索结果进行排列，PageRank揭示于名称为“链接数据库中的节点排列方法(Method for noderanking in a linked database)”的第6,285,999号美国专利中，该专利的全文及其全部教示内容以引用方式并入本文中。应注意，所述搜索结果通常与Stanford University(斯坦福大学)有关。In the example shown in FIG. 6 , thepersonalization controller 608 is set all the way to itsminimum value 612 . Therefore, the corresponding search results are ranked according to their relevance to the query and information irrelevant to the query (such as PageRank), which is disclosed in the method named "Method for noderanking in a linked database". )", which is hereby incorporated by reference in its entirety and in its entirety for its teachings. It should be noted that the search results are generally related to Stanford University.

现在参见图7，图中显示在用户已将滑标610调整至一位于最小值与最大值之间大约处于中点的中间位置后对搜索结果602的排列。这时，各个文档604得到重新排列，且与用户的兴趣更为相关的那些(前面标识为“Computers(General))”具有更高的次序。为了使用户辨别与他的兴趣更为相关的文档，那些文档由一紧挨着链接文本615的个性化图标618指示。注意到，在滑标610处于中间位置的情况下，大约每隔一个文档便为一个性化结果，尽管并非必需如此。进一步注意到，片断619文本具有与计算机相关的文本(例如“hacking(非法闯入)”)，即使该用语并未包括于原始搜索查询中。Referring now to FIG. 7, there is shown the ranking ofsearch results 602 after the user has adjustedslider 610 to a position approximately midway between the minimum and maximum values. At this point, thevarious documents 604 are rearranged, with those that are more relevant to the user's interests (identified earlier as "Computers (General)") having a higher order. In order for the user to identify documents that are more relevant to his interests, those documents are indicated by apersonalized icon 618 next to thelink text 615 . Note that withslider 610 in the middle position, approximately every other document is a personalized result, although this is not required. Note further that thesnippet 619 text has computer-related text (eg "hacking") even though that term was not included in the original search query.

现在参见图8，图中显示在用户已将滑标610调整至其最大位置614后搜索结果602的排列。首先，在此实例中，搜索结果602中的所有文档604均反映用户的个性化配置文件并以个性化图标618来注解。其次，所述搜索结果中的文档与计算机及Stanford大学相关，从而反映用户对计算机的兴趣，因为其与查询“Stanford”相关。例如，前两个结果与在Stanford的“非法闯入”相关；接下来的三个文档具体与Stanford的各种计算机科学项目相关。因此，第三个文档是针对Stanford的Artificial Intelligence Laboratory(人工智能实验室)，第四个文档是针对Stanford的Center for the Study of Language and Information(语言与信息研究中心)，而第五个文档是针对Stanford的Database Group(数据库小组)。其余文档同样与用户的“Computers”兴趣相关。进一步注意到，此前在当不应用个性化(图6)及当只应用中等个性化(图7)时作为第一(例如最相关的)结果的与Stanford大学的主页的链接不再出现于前八个结果中。Referring now to FIG. 8 , there is shown the arrangement ofsearch results 602 after the user has adjustedslider 610 to itsmaximum position 614 . First, alldocuments 604 insearch results 602 reflect the user's personalized profile and are annotated with apersonalized icon 618 in this example. Second, the documents in the search results are related to computers and Stanford University, reflecting the user's interest in computers because it is relevant to the query "Stanford." For example, the first two results are related to "break and enter" at Stanford; the next three documents are specifically related to various computer science programs at Stanford. So the third document is for Stanford's Artificial Intelligence Laboratory, the fourth document is for Stanford's Center for the Study of Language and Information, and the fifth document is For Stanford's Database Group (database group). The remaining documents are similarly related to the user's "Computers" interest. Note further that the link to Stanford University's home page, which was previously the first (e.g., most relevant) result when no personalization is applied (FIG. 6) and when only medium personalization is applied (FIG. 7), no longer appears in the top of eight results.

在图6-8所示实例中，用户可实时地对文档的排列加以动态调整，而搜索引擎无需重新计算搜索结果。此进一步有助于用户能够探究所述搜索结果。In the example shown in Figures 6-8, the user can dynamically adjust the arrangement of the documents in real time, and the search engine does not need to recalculate the search results. This further facilitates the user's ability to explore the search results.

如图6-8所示，一Edit Profile(编辑配置文件)链接616使用户能够随时对他的用户配置文件的兴趣进行编辑。图9图解说明用户已返回至图2中所示的主题目录页面200、删除了他现有的Computers主题兴趣、并创建了“Music(音乐)”这一不同的兴趣。然后，用户重新选择Start Searching(开始搜索)按钮208。搜索引擎提供经更新的一组搜索结果，所述经更新的一组搜索结果的文档将与以前的相同，因为尚未改变所述查询术语。然而，对于最后一次搜索，用户已将控制器610定位于完全个性化处，且因此所述搜索引擎应用此设置并根据现有的(及新的)个性化配置文件(此时针对“Music(音乐)”)对文档进行排列。图10图解说明此在所述搜索结果中所产生的巨大差异。此处，顶部的六个结果均与音乐及艺术相关，而先前关于计算机的结果均不处于所述搜索结果列表的顶部。此表明用户的个性化配置文件可如何显著地改变对给定的一组文档的排列。As shown in Figures 6-8, an Edit Profile (edit configuration file)link 616 enables the user to edit his user profile interests at any time. Figure 9 illustrates that the user has returned to thesubject category page 200 shown in Figure 2, deleted his existing Computers subject interest, and created a different interest of "Music". Then, the user reselects the Start Searching (start searching)button 208. The search engine provides an updated set of search results whose documents will be the same as before because the query terms have not been changed. However, for the last search, the user has positioned thecontroller 610 at full personalization, and so the search engine applies this setting and according to the existing (and new) personalization profile (this time for "Music( music)") to arrange the documents. Figure 10 illustrates the huge difference this makes in the search results. Here, the top six results are all related to music and art, while none of the previous results about computers were at the top of the search result list. This demonstrates how a user's personalized profile can significantly alter the ranking of a given set of documents.

在本发明的一实施例中，对搜索结果的个性化是由下述方法实现的。搜索引擎设计领域的技术人员可容易理解其中可采用此方法的搜索引擎的基本结构及其操作。In an embodiment of the present invention, the personalization of search results is realized by the following method. Those skilled in the art of search engine design can readily understand the basic structure of a search engine in which this method can be employed and its operation.

1)当用户选择主题(类别)时，所述主题(类别)被添加至一包含他的兴趣IN(其中IN为“interests(兴趣)”的缩写)的列表的跟踪文件中。所述兴趣可表示成文本符记(例如“computers(计算机)”)或表示成加入目录中的索引，其中每一主题/副主题均具有一索引号。当用户访问所述站点时，系统自所述跟踪文件检索此兴趣列表。或者，所述站点可提供一使用户在其中登记一用户名及口令的用户登记页面，并将用户的兴趣在搜索引擎站点处存储于一数据库中，并响应于用户登录来对其进行检索。1) When a user selects a topic (category), said topic (category) is added to a tracking file containing a list of his interests IN (where IN is an abbreviation for "interests"). The interest can be represented as a text notation (such as "computers") or as an index added to a table of contents, where each topic/subtopic has an index number. The system retrieves this interest list from the tracking file when the user visits the site. Alternatively, the site may provide a user registration page where the user registers a username and password, and store the user's interests in a database at the search engine site and retrieve them in response to user login.

2)每当用户发出一搜索查询Q时，便将所述列表IN发送至一前端网络服务器。2) Whenever a user issues a search query Q, the list IN is sent to a front-end web server.

3)然后，所述前端网络服务器将两个请求(Q，空)及(Q，IN)发给一后端搜索引擎，所述后端搜索引擎向所述前端网络服务器返回两个别为R_base(常规的)及R_in(个性化的)结果集合。3) Then, the front-end web server sends two requests (Q, empty) and (Q, IN) to a back-end search engine, and the back-end search engine returns two R_base( regular) and R_in (personalized) result sets.

4)所述基础查询(Q，空)与在搜索引擎上在不施加任何个性化情况下对Q进行的搜索相同，因而R_base为所述搜索引擎通常将响应于所述查询而返回的结果集合。如上文所述，根据其与所述查询的相关性及例如PageRank等任何其它因素来相应地对所述搜索结果进行排列，PageRank揭示于名称为“链接数据库中的节点排列方法(Method for node ranking in a linked database)的第6,285,999号美国专利中，该专利的全文及其全部教示内容以引用方式并入本文中。4) The base query (Q, empty) is the same as a search on Q performed on a search engine without any personalization applied, so R_base is the set of results that the search engine would normally return in response to the query . As mentioned above, the search results are ranked accordingly according to their relevance to the query and any other factors such as PageRank disclosed in the article titled "Method for node ranking in link database". 6,285,999 in a linked database), which is hereby incorporated by reference in its entirety and for its entire teachings.

5)个性化的查询(Q，IN)包含用户的兴趣列表。所述搜索引擎遵循一类似于常规查询的处理控制流程，但其在记分功能期间激活一个性化组件，以“提升”与所述兴趣IN集合相关的文档。5) The personalized query (Q, IN) contains the user's interest list. The search engine follows a process control flow similar to a regular query, but it activates a personalization component during the scoring function to "promote" documents relevant to the IN set of interest.

在本发明的一实施例中，使用一组下列形式的映射(所述目录中的每一主题/副主题各一个映射)来实施这些提升：In one embodiment of the invention, these promotions are implemented using a set of maps (one for each topic/subtopic in the catalog) of the form:

{站点-＞提升}{site->boost}

因而例如，类别“Health(健康)”具有如下形式的映射So for example, the category "Health" has a mapping of the form

{站点-＞提升}{site->boost}

其包含下列输入：It contains the following inputs:

nih.gov-＞5.8nih.gov->5.8

cdc.gov--＞7.9cdc.gov-->7.9

med.stanford.edu-＞3.5med.stanford.edu->3.5

其中nih.gov，cdc.gov及med.Stanford.edu为已确定出(以手动或自动方式)与主题“Health”相关的各个站点。换句话说，对于所述目录中的每一主题，均存在一组已确定出与所述主题相关的站点，且对于这些站点中的每一站点，均界定一提升值。Among them, nih.gov, cdc.gov and med.Stanford.edu are the sites that have been determined (manually or automatically) related to the topic "Health". In other words, for each topic in the directory there is a set of sites that have been determined to be relevant to the topic, and for each of these sites a lift value is defined.

通常按下述方式来确定所述主题目录中所列出的站点的提升：The promotion of a site listed in the subject directory is generally determined as follows:

a)产生一“站点图”，其中所述图的各节点为站点(基本上为同一主机上的页面)，并根据来自一个站点的链接至另一主机上的页面的的页面数量来对各节点之间的边缘进行加权。可使用此同一类型的图来计算所有主题提升映射。a) Create a "site graph" where the nodes of the graph are sites (essentially pages on the same host) and rank each Edges between nodes are weighted. All topic lift maps can be calculated using this same type of graph.

b)对于所述目录中的每一主题，例如“Health”，选择若干个其主页列示于OpenDirectory(开放式目录)中的站点作为“起始”站点S0。例如，对于一所大学，如Stanford，起始站点可选择成任一以.stanford.edu结尾的站点。b) For each topic in the directory, eg "Health", select several sites whose home pages are listed in OpenDirectory (Open Directory) as "starting" sites S0. For example, for a university such as Stanford, the starting site can be any site ending in .stanford.edu.

c)分两遍进行计算：c) Calculate in two passes:

i)首先，识别一组由S0中的那些站点所大量链接的站点S1，其中根据S1中的每一站点如何大量地被S0中的站点所链接来为其指配一权重。i) First, identify a set of sites S1 that are heavily linked by those sites in S0, where each site in S1 is assigned a weight according to how heavily linked by sites in S0.

ii)然后，识别那些由S1中的那些站点所大量链接的站点S2，并如同S1一样对其进行加权。ii) Then, identify those sites S2 that are heavily linked by those in S1 and weight them like S1.

d)以其所指配到的权重来提升S2中的站点。d) Promote the station in S2 by its assigned weight.

当个性化查询具有一兴趣集合(例如{Health(健康)，Music(音乐)})时，则所述个性化组件查阅各自的映射，且随后对于处于这些映射中的查询结果(即文档)，所述个性化组件提升这些查询结果的IR得分，以便其将获得更高的排列次序。所述IR得分为所述搜索引擎为每一文档提供的信息检索得分，并可随文档相关性及例如PageRank等其它因素而变化。When a personalized query has a collection of interests (e.g., {Health (health), Music (music)}), then the personalization component consults the respective mappings, and then for the query results (i.e., documents) in these mappings, The personalization component boosts the IR score of these query results so that it will get a higher rank. The IR score is the information retrieval score provided by the search engine for each document and may vary with document relevance and other factors such as PageRank.

例如，假如http://www.med.stanford.edu/research/为查询Q及包含“Health”的IN的结果之一，则因所述页面位于站点med.stanford.edu上，所述结果的IR得分将得到一3.5倍的倍增提升。对于多个兴趣而言，将各个对应映射中的值一起相乘以得到一组合效果。一种用于为所述目录提升映射中的站点确定所述提升值的实施方式的进一步细节阐述于第10/646,331号美国专利申请案中，该专利申请案以引用方式并入本文中。For example, if http://www.med.stanford.edu/research/ is one of the results for query Q and IN containing "Health", then since the page is on the site med.stanford.edu, the The IR score will get a 3.5x multiplier. For multiple interests, the values in each corresponding map are multiplied together to obtain a combined effect. Further details of an embodiment for determining the boost value for a site in the directory boost map are set forth in US patent application Ser. No. 10/646,331, which is incorporated herein by reference.

6)所述前端网络服务器接收R_base及R_in。R_base代表在个性化控制器608设定为0(或其最小位置612)处时用户所看到的结果602。R_in代表在所述滑标处于其最大位置614(即10)时用户所看到的结果602。在一实施例中，所述控制器具有11个位置，但也可使用任何数量的位置。对显示给用户的介于0与10之间的排列进行内插；这意味着，借助R_base及R_in，所述个性化组件通过使每一单独结果沿一自其在控制器608处于其最小位置时的位置至在控制器608处于其最大位置时的位置的虚拟路径移动来构造中间排列(R_1...R_9)。6) The front-end network server receives R_base and R_in. R_base represents theresult 602 that the user sees when thepersonalization control 608 is set to 0 (or its minimum position 612). R_in represents theresult 602 seen by the user when the slider is at its maximum position 614 (ie 10). In one embodiment, the controller has 11 positions, but any number of positions could be used. Interpolate the permutations displayed to the user between 0 and 10; this means that, by means of R_base and R_in, the personalization component is at its minimum position in thecontroller 608 by making each individual result along a The intermediate arrays (R_1 .

7)如何向用户提供滑标效果：7) How to provide users with a sliding effect:

所述前端网络服务器将R_in中不处于R_base中的结果附加至R_base，以获得一称作R_all的关于所有唯一结果(在R_base或R_in中)的列表。然后，所述前端网络服务器为R_all中的结果指配连续的识别号以充当识别符。然后，所述前端网络服务器向用户的浏览器发送两个项：The front-end web server appends the results in R_in that are not in R_base to R_base to obtain a list of all unique results (either in R_base or R_in) called R_all. The front-end web server then assigns consecutive identification numbers to the results in R_all to serve as identifiers. The front-end web server then sends two items to the user's browser:

a)R_all中每一结果的映射{识别号-＞结果}，对于每一文档其均在所述结果中包含其识别号；及a) a mapping {ID -> Result} for each result in R_all, which includes, for each document, its ID in the result; and

b)N个向量，其标记用户应看到(以结果识别号表示)的排列，其中N为滑标的位置数量，且每一向量均代表根据个性化程度的结果排序。在此实例中，将具有11个向量对应于控制器608的位置0至10。在一实施例中，向量0为[0，1，2，3，4，5，6，7，8，9，10]，其反映对指配有以R_0开始的本地ID的结果的非个性化排列。向量10为完全个性化的结果R_in中的排序；举例而言，其可能是[15，17，5，0，1，2，3，4，8，9，10]，其中文档15及17(例如)为其IR得分根据映射图中的提升值及与控制器608的位置相关联的值得到提升的主题目录的相关映射中的文档。将所述N个排列向量传输至客户端便会使所述客户端浏览器能够快速地重新计算所述排列，而不是使服务器承担所述计算。b) N vectors, which mark the arrangement that the user should see (represented by the result identification number), where N is the number of positions of the slider, and each vector represents the ranking of results according to the degree of personalization. In this example, there will be 11 vectors corresponding to positions 0-10 of thecontroller 608 . In one embodiment, vector 0 is [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10], which reflects the impersonality of the results assigned with local IDs starting with R_0 arrangement. Vector 10 is the ranking in the fully personalized result R_in; for example, it might be [15, 17, 5, 0, 1, 2, 3, 4, 8, 9, 10], where documents 15 and 17 ( For example) a document in a relevant map of a subject directory whose IR score is promoted according to the promotion value in the map and the value associated with the location of thecontroller 608 . Transmitting the N permutation vectors to the client enables the client browser to quickly recalculate the permutation, rather than burdening the server with the computation.

8)最后，所述结果页面中的Javascript代码使用户的浏览器响应于用户改变控制器608的位置而使用所述11个位置排列向量中适当的一个来对所述搜索结果进行重新排序。上述步骤6-8的一实施例的进一步细节阐述于第10/641,489号美国申请案中，该申请案以引用方式并入本文中。8) Finally, the Javascript code in the results page causes the user's browser to reorder the search results using the appropriate one of the 11 position ordering vectors in response to the user changing the position of thecontrol 608 . Further details of one example of steps 6-8 above are set forth in US application Ser. No. 10/641,489, which is incorporated herein by reference.

上文已参照一可能的实施例具体地详细阐述了本发明。所属领域的技术人员将了解，本发明也可实施为其它实施例。首先，对各组件的具体命名、大写的用语、属性、数据结构、或任何其它编程或结构方面均不是强制性的或有特殊意义的，且用于构建本发明或其特征的机理可具有不同的名称、格式或协议。此外，所述系统可如上文所述通过一硬件与软件的组合构建而成，或完全构建于硬件元件中。此外，本文中所述的在各种系统组件之间进行的特定的功能划分仅为实例性的，而不是强制性的；由单个系统组件所实施的功能也可由多个组件实施，而由多个组件所实施的功能也可改为由单个组件实施。The invention has been explained in detail above with reference to a possible embodiment. Those skilled in the art will appreciate that the invention can be practiced in other embodiments as well. First, the specific naming, capitalization of terms, attributes, data structures, or any other programming or structural aspects of each component are not mandatory or have special meaning, and the mechanisms used to construct the invention or its features may have different The name, format or protocol of the . Furthermore, the system may be implemented as described above by a combination of hardware and software, or entirely in hardware elements. In addition, the specific functional division among various system components described herein is only exemplary, not mandatory; functions implemented by a single system component may also be implemented by multiple components, and multiple components may be implemented by multiple components. Functions implemented by individual components may instead be implemented by a single component.

上述说明的某些部分以对信息的操作的算法及符号表示形式来提供本发明的特征。这些算法说明及表示形式是数据处理领域的技术人员用于最有效地将其工作的实质传达给所属领域的其它技术人员的手段。这些操作在以功能或逻辑方式加以阐述时应理解为由计算机程序实施。此外，还已证明，有时将这些操作安排称作模块或按功能名称来称谓是方便的，此并不失其一般性。Certain portions of the above description present features of the invention in terms of algorithms and symbolic representations of operations on information. These algorithmic descriptions and representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work to others skilled in the art. These operations, when described in a functional or logical manner, should be understood as being implemented by computer programs. Furthermore, it has also proven convenient at times, to refer to these arrangements of operations as modules or by functional names, without loss of generality.

除非根据上述说明显而易见地另有具体规定，否则应了解，在本说明的通篇中，利用诸如“计算”或“确定”或“识别”或类似用语进行的论述是指一计算机系统或类似电子计算装置所进行的步骤及过程，所述计算机系统或类似电子计算装置对在计算机系统存储器或寄存器或其它此类信息存储器、传输或显示装置中被表示成物理(电子)量的数据进行处理及变换。Unless it is obvious from the above description that it is specifically stated otherwise, it should be understood that throughout this specification, discussions using terms such as "calculate" or "determine" or "identify" or similar terms refer to a computer system or similar electronic steps and processes performed by a computing device that processes and processes data represented as physical (electronic) quantities in computer system memory or registers or other such information storage, transmission or display devices transform.

本发明的某些方面包括在本文中描述成算法形式的过程步骤及指令。应注意，本发明的过程步骤及指令可收录于软件、固件或硬件中，且当收录于软件中时，可下载驻存于实时网络操作系统所使用的不同平台上并可自所述不同平台运行。Certain aspects of the invention include process steps and instructions described herein in algorithmic form. It should be noted that the process steps and instructions of the present invention may be embodied in software, firmware, or hardware, and when embodied in software, may be downloaded, resident, and accessible from different platforms used by real-time network operating systems. run.

本发明还涉及一种用于实施本文中的操作的设备。此设备可针对所需用途专门构造，或者其可包括一通用计算机，所述通用计算机由一存储于一可由计算机存取的计算可读媒体上的计算机程序有选择地启动或重新配置。此种计算机程序可存储于一计算机可读存储媒体中，例如(但不限于)任何类型的磁盘，包括软盘、光盘、CD-ROM、磁-光盘、只读存储器(ROM)、随机存取存储器(RAM)、EPROM、EEPROM、磁卡或光卡，或适于存储电子指令并分别耦接至一计算机系统总线的任何类型的媒体。集成电路设计及视频编解码器领域的技术人员应了解，本发明可容易地制作于基于上述功能及结构说明的各种类型的集成电路(包括应用专用集成电路(ASIC))中。另外，本发明可包含于各种类型的视频编码装置内。The present invention also relates to an apparatus for carrying out the operations herein. This apparatus may be specially constructed for the required purposes, or it may comprise a general purpose computer selectively activated or reconfigured by a computer program stored on a computer-readable medium accessible by the computer. Such a computer program may be stored on a computer readable storage medium such as, but not limited to, any type of disk, including floppy disk, compact disk, CD-ROM, magneto-optical disk, read-only memory (ROM), random-access memory (RAM), EPROM, EEPROM, magnetic or optical card, or any type of media suitable for storing electronic instructions and respectively coupled to a computer system bus. Those skilled in the art of integrated circuit design and video codecs should understand that the present invention can be easily fabricated in various types of integrated circuits (including application specific integrated circuits (ASICs)) based on the above functional and structural descriptions. Additionally, the present invention can be included within various types of video encoding devices.

本文中所提供的算法及操作并非与任一特定计算机或其它设备内在地相关。各种通用系统也可与根据本文教示内容的程序一起使用，或者可证明能方便地构造更专门的设备来实施所需方法步骤。所属领域的技术人员将易知各种此类系统所需的结构以及等效变化形式。另外，本发明未参照任一特定编程语言加以阐述。应了解，可使用各种编程语言来实施本文中所述的本发明教示内容，且特定语言的任何参考均旨在揭示本发明的实现方式及最佳模式。The algorithms and operations presented herein are not inherently related to any particular computer or other device. Various general-purpose systems may also be used with programs in accordance with the teachings herein, or it may prove convenient to construct more specialized apparatus to carry out the required method steps. The required structure, and equivalent variations, for a variety of such systems will be readily apparent to those skilled in the art. In addition, the invention has not been described with reference to any particular programming language. It should be appreciated that a variety of programming languages can be used to implement the teachings of the invention described herein, and any reference to a specific language is intended to reveal the best mode and implementation of the invention.

最后，应注意，本说明书中所使用的语言原则上是出于可读性和指导性目的而选取，且可能不是选取用来描绘或限制本发明标的物。因此，本发明揭示内容旨在例示而非限制本发明的范围。Finally, it should be noted that the language used in this specification has been chosen for readability and instructional purposes in principle, and may not have been chosen to delineate or limit the inventive subject matter. Accordingly, the present disclosure is intended to be illustrative, not limiting, of the scope of the present invention.

Claims (36)

Translated fromChinese

1、一种用于根据一用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：CLAIMS 1. A computer-implemented method for personalizing results of a search query based on a user, comprising:检索一组与所述用户的所述搜索查询相关的文档；retrieving a set of documents relevant to said search query of said user;响应于所述用户的一兴趣配置文件及所述兴趣配置文件对所述组文档的一第一排列的一影响程度来确定所述排列；及determining the ranking in response to an interest profile of the user and a degree of influence of the interest profile on a first ranking of the set of documents; and将所述经排列的文档呈现给所述用户。The arranged documents are presented to the user.2、如权利要求1所述的方法，其中呈现所述经排列的文档进一步包括：2. The method of claim 1, wherein presenting the arranged documents further comprises:紧挨着与一包含于所述用户的所述兴趣配置文件中的兴趣相关联的每一文档显示一个性化图标。A personalized icon is displayed next to each document associated with an interest contained in the interest profile of the user.3、如权利要求1所述的方法，其进一步包括：3. The method of claim 1, further comprising:自所述用户接收一所述用户的兴趣配置文件将对所述经排列的文档产生的一影响程度的输入；receiving from the user an input of a degree of influence that the user's interest profile will have on the ranked document;响应于所述影响程度来确定所述组文档的一第二排列；及determining a second arrangement of the set of documents in response to the degree of influence; and以所述第二排列将所述文档呈现给所述用户。The document is presented to the user in the second arrangement.4、如权利要求3所述的方法，其中在不检索与所述查询相关的一第二组文档的情况下实施对所述第二排列的确定。4. The method of claim 3, wherein determining the second ranking is performed without retrieving a second set of documents relevant to the query.5、如权利要求3所述的方法，其中在接收所述影响程度的所述用户输入前确定所述第二排列。5. The method of claim 3, wherein said second ranking is determined prior to receiving said user input of said degree of influence.6、如权利要求3所述的方法，其中与确定所述第一排列同时确定所述第二排列。6. The method of claim 3, wherein the second permutation is determined concurrently with the first permutation.7、如权利要求3所述的方法，其中所述控制程度的所述用户输入为一控制器的一图形表示中一位置变化。7. The method of claim 3, wherein said user input of said degree of control is a position change in a graphical representation of a controller.8、如权利要求7所述的方法，其中所述控制器为一滑标。8. The method of claim 7, wherein the controller is a slider.9、如权利要求7所述的方法，其中所述控制器为一刻度盘。9. The method of claim 7, wherein the controller is a dial.10、如权利要求7所述的方法，其中所述控制器为无线电按钮。10. The method of claim 7, wherein the controller is a radio button.11、如权利要求1所述的方法，其中确定一第一排列进一步包括：11. The method of claim 1, wherein determining a first arrangement further comprises:自一跟踪文件检索所述用户的所述兴趣配置文件。The interest profile for the user is retrieved from a tracking file.12、如权利要求1所述的方法，其中确定一第一排列进一步包括：12. The method of claim 1, wherein determining a first arrangement further comprises:自一以通信方式耦接至一搜索引擎的数据库检索所述用户的所述兴趣配置文件。The interest profile of the user is retrieved from a database communicatively coupled to a search engine.13、如权利要求1所述的方法，其中所述用户的所述兴趣配置文件包括一组文本符记，每一文本符记对应于一兴趣主题。13. The method of claim 1, wherein the interest profile of the user includes a set of textual tokens, each textual token corresponding to a topic of interest.14、如权利要求1所述的方法，其中所述用户的所述兴趣配置文件包括一组主题索引号，每一索引号对应于一兴趣主题。14. The method of claim 1, wherein the interest profile of the user includes a set of topic index numbers, each index number corresponding to a topic of interest.15、如权利要求1所述的方法，其中确定一第一排列包括：15. The method of claim 1, wherein determining a first arrangement comprises:响应于与一包含于所述用户的所述兴趣配置文件中的兴趣相关联的所述文档，用一与所述文档相关联的提升值来按比例缩放所述组中的至少一个文档的信息检索得分。responsive to said document associated with an interest contained in said interest profile of said user, scaling information of at least one document in said set with a boost value associated with said document Retrieve score.16、如权利要求1所述的方法，其中每一文档均具有一信息检索得分，且确定一第一排列包括：16. The method of claim 1, wherein each document has an information retrieval score, and determining a first ranking comprises:对于每一与一包含于所述用户的所述兴趣配置文件中的主题相关联的文档，均为所述文档确定一提升值，并用所述提升值及所述影响程度来按比例缩放所述文档的一信息检索得分；及For each document associated with a topic contained in the interest profile of the user, a lift value is determined for the document, and the boost value and the degree of influence are used to scale the an information retrieval score for the document; and根据其信息检索得分来对所述组中的所述文档进行排列。The documents in the group are arranged according to their information retrieval scores.17、如权利要求1所述的方法，其中确定一第一排列包括：17. The method of claim 1, wherein determining a first arrangement comprises:对于所述组文档中的至少一个文档：For at least one document in the set of documents:确定一包含所述文档的网站是否与一包含于所述用户的所述兴趣配置文件中的主题相关联；及determining whether a website containing the document is associated with a topic contained in the user's interest profile; and用与包含所述文档的所述网站相关联的所述提升值来按比例缩放所述文档的一信息检索得分。An information retrieval score for the document is scaled by the boost value associated with the website containing the document.18、如权利要求1所述的方法，其中确定一第一排列包括：18. The method of claim 1, wherein determining a first arrangement comprises:对于所述组文档中的至少一个文档：For at least one document in the set of documents:确定所述文档是否包含于一由一第二网站链接到的第一网站中，所述第二网站与一包含于所述用户的所述兴趣配置文件中的主题相关联；及determining whether the document is contained in a first website linked to by a second website associated with a topic contained in the interest profile of the user; and用一与所述第一网站相关联的加权值来按比例缩放所述文档的一信息检索得分。An information retrieval score for the document is scaled with a weight value associated with the first website.19、如权利要求18所述的方法，其中所述加权值是基于一自所述第二网站至所述第一网站的链接的数量。19. The method of claim 18, wherein said weighting value is based on a number of links from said second website to said first website.20、如权利要求1所述的方法，其中确定一第一排列包括：20. The method of claim 1, wherein determining a first arrangement comprises:对于一与复数个包含于所述用户的所述兴趣配置文件中的主题相关联的文档：For a document associated with a plurality of topics contained in said interest profile of said user:确定一包含所述文档的网站；identifying a website containing the document;对于每一与所述网站相关联的主题，均为一网站确定一提升值；及determining a lift value for a website for each theme associated with said website; and作为所述提升值的一功能，为所述文档确定一最终提升值；及determining a final boost value for the document as a function of the boost value; and用所述最终提升值来按比例缩放所述文档的所述信息检索得分。The information retrieval score of the document is scaled by the final boost value.21、如权利要求1所述的方法，其中确定一第一排列包括：21. The method of claim 1, wherein determining a first arrangement comprises:从一包括复数个主题的主题目录中为至少一个文档确定一提升值，每一主题均与至少一个网站相关联，每一网站均具有一提升值，其中所述用户的所述兴趣配置文件包含所述主题目录中的至少一个主题；及Determining a lift value for at least one document from a subject directory comprising a plurality of topics, each topic associated with at least one website, each website having a lift value, wherein said interest profile of said user includes at least one subject in said subject catalog; and根据其提升值来按比例缩放所述至少一个文档的一信息检索得分。An information retrieval score for the at least one document is scaled according to its boost value.22、一种用于根据一用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：22. A computer-implemented method for personalizing results of a search query based on a user, comprising:检索一组与所述用户的所述搜索查询相关的文档；retrieving a set of documents relevant to said search query of said user;独立于所述用户的一兴趣配置文件来确定所述组文档的一第一排列；determining a first ranking of the set of documents independently of an interest profile of the user;将所述经排列的文档呈现给所述用户；presenting the arranged documents to the user;自所述用户接收一输入，所述输入提供所述用户的兴趣配置文件将对所述经排列的文档产生的一影响程度；及receiving an input from the user providing a degree of influence that the user's interest profile will have on the ranked document; and根据所述影响程度以一第二排列将所述文档呈现给所述用户。The documents are presented to the user in a second arrangement according to the degree of influence.23、一种用于根据一用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：23. A computer-implemented method for personalizing results of a search query based on a user, comprising:检索一组与所述用户的所述搜索查询相关的文档；retrieving a set of documents relevant to said search query of said user;根据所述用户的一兴趣配置文件的一最小影响程度来确定所述组文档的一第一排列；determining a first ranking of the set of documents based on a minimum degree of influence of an interest profile of the user;根据所述用户的所述兴趣配置文件的一最大影响程度来确定所述组文档的一第二排列；determining a second ranking of the set of documents based on a maximum degree of influence of the interest profile of the user;将对所述文档的所述第一排列呈现给所述用户；presenting the first arrangement of the documents to the user;接收一用户输入，所述用户输入提供一所述用户的兴趣配置文件的一最大影响程度的选择；receiving a user input providing a selection of a maximum degree of influence for the user's interest profile;将所述文档的所述第二排列呈现给所述用户。The second arrangement of the documents is presented to the user.24、一种用于根据用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：24. A computer-implemented method for personalizing results of a search query based on a user, comprising:检索一组与所述用户的所述搜索查询相关的文档；retrieving a set of documents relevant to said search query of said user;根据所述用户的所述兴趣配置文件的一最大影响程度来确定所述组文档的一第一排列；determining a first ranking of the set of documents based on a maximum degree of influence of the interest profile of the user;根据所述用户的一兴趣配置文件的一最小影响程度来确定所述组文档的一第二排列；determining a second ranking of the set of documents based on a minimum degree of influence of an interest profile of the user;将所述文档的所述第一排列呈现给所述用户；presenting the first arrangement of the documents to the user;接收一用户输入，所述用户输入提供一所述用户的兴趣配置文件的一最小影响程度的选择；receiving a user input providing a minimally influential selection of the user's interest profile;将所述文档的所述第二排列呈现给所述用户。The second arrangement of the documents is presented to the user.25、一种用于根据一用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：25. A computer-implemented method for personalizing results of a search query based on a user, comprising:检索一组与所述用户的所述搜索查询相关的文档；retrieving a set of documents relevant to said search query of said user;根据所述用户的一兴趣配置文件的一最小影响程度来确定所述组文档的一第一排列；determining a first ranking of the set of documents based on a minimum degree of influence of an interest profile of the user;根据所述用户的所述兴趣配置文件的一最大影响程度来确定所述组文档的一第二排列；determining a second ranking of the set of documents based on a maximum degree of influence of the interest profile of the user;将对所述文档的所述第一排列呈现给所述用户；presenting the first arrangement of the documents to the user;接收一用户输入，所述用户输入提供所述用户的兴趣配置文件的一影响程度；receiving a user input providing a degree of influence of the user's interest profile;根据相对于所述最小及最大影响程度的所述所接收到的影响程度通过在所述第一排列与所述第二排列之间进行内插来确定所述文档的一第三排列。A third ranking of the document is determined by interpolating between the first ranking and the second ranking based on the received influence levels relative to the minimum and maximum influence degrees.26、一种用于根据一用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：26. A computer-implemented method for personalizing results of a search query based on a user, comprising:检索一组与所述用户的所述搜索查询相关的文档；retrieving a set of documents relevant to said search query of said user;响应于所述用户的一兴趣配置文件及所述兴趣配置文件对所述组文档的一第一排列的一影响程度来确定所述组文档的所述排列，其中所述兴趣配置文件包括所述用户感兴趣的至少一个主题；determining the ranking of the set of documents in response to an interest profile of the user and a degree of influence of the interest profile on a first ranking of the set of documents, wherein the interest profile includes the at least one topic of interest to the user;将所述经排列的文档呈现给所述用户；presenting the arranged documents to the user;自所述用户接收一对兴趣配置文件的修改，所述改变包括至少如下中的一种：a)至少一个主题添加至所述兴趣配置文件，及b)至少一个主题自所述兴趣配置文件中删除；receiving a modification to an interest profile from the user, the change comprising at least one of: a) at least one topic added to the interest profile, and b) at least one topic removed from the interest profile delete;响应于所述经修改的兴趣配置文件来确定所述组文档的一第二排列；及determining a second arrangement of the set of documents in response to the modified interest profile; and以所述第二排列将所述文档呈现给所述用户。The document is presented to the user in the second arrangement.27、一种用于根据一用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：27. A computer-implemented method for personalizing results of a search query based on a user, comprising:在一服务器处：At a server:检索一组与一用户的一查询相关的文档；retrieving a set of documents relevant to a query by a user;根据所述用户的一兴趣配置文件的一最小影响程度来确定所述组文档的一第一排列；determining a first ranking of the set of documents based on a minimum degree of influence of an interest profile of the user;根据所述用户的所述兴趣配置文件的一最大影响程度来确定所述组文档的一第二排列；determining a second ranking of the set of documents based on a maximum degree of influence of the interest profile of the user;确定复数个排列向量，包括一描述所述组文档的所述第一排列的第一排列向量、一描述所述组文档的所述第二排列的第二排列向量、及至少一个中间排列向量，所述中间排列向量描述一内插于所述第一与第二排列之间的所述组文档的第三排列；determining a plurality of permutation vectors comprising a first permutation vector describing said first permutation of said set of documents, a second permutation vector describing said second permutation of said set of documents, and at least one intermediate permutation vector, said intermediate permutation vector describes a third permutation of said set of documents interpolated between said first and second permutations;在一客户端处：At a client:根据所述第一排列向量将所述文档的所述第一排列呈现给所述用户；presenting the first permutation of the document to the user according to the first permutation vector;接收一用户输入，所述用户输入提供所述用户的兴趣配置文件的一影响程度；receiving a user input providing a degree of influence of the user's interest profile;从所述复数个排列向量中选择一对应于所述所接收到的所述影响程度的用户输入的排列向量；及selecting a permutation vector corresponding to the received user input of the degree of influence from the plurality of permutation vectors; and根据所述所选定的排列向量来呈现所述组文档。The set of documents is presented according to the selected permutation vectors.28、一种用于根据一用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：28. A computer-implemented method for personalizing results of a search query based on a user, comprising:检索一组与所述用户的所述搜索查询相关的文档；retrieving a set of documents relevant to said search query of said user;根据一组由所述用户选择为代表所述用户的兴趣的主题及所述所选主题的一影响程度来对所述组文档进行排列；及ranking said set of documents according to a set of topics selected by said user to represent said user's interests and a degree of influence of said selected topics; and将所述经排列的文档呈现给所述用户。The arranged documents are presented to the user.29、一种用于根据一用户将一搜索查询的结果个性化的由计算机实施的方法，其包括：29. A computer-implemented method for personalizing results of a search query based on a user, comprising:检索一组与所述用户的所述搜索查询相关的文档，每一文档均具有一信息检索得分；retrieving a set of documents relevant to said search query by said user, each document having an information retrieval score;根据一指配给一与所述文档相关的网站的提升值来提升所述组文档中至少一个文档的所述信息检索得分，其中所述网站与一组由所述用户选择为代表所述用户的兴趣的主题中的至少一个主题相关联；boosting the information retrieval score for at least one document in the set of documents according to a boost value assigned to a website associated with the document, wherein the website is associated with a set of websites selected by the user to represent the user associated with at least one of the topics of interest;根据其信息检索得分来对所述组文档进行排列，包括具有经提升的信息检索得分及未经提升的信息检索得分的文档；及ranking the set of documents according to their information retrieval scores, including documents with boosted information retrieval scores and non-boosted information retrieval scores; and将所述经排列的文档呈现给所述用户。The arranged documents are presented to the user.30、一种用于一搜索引擎的用户界面，其包括：30. A user interface for a search engine comprising:一包括复数个主题的主题目录，其包括可由一用户选择的主题以指示一所述主题中用于界定所述用户的一兴趣配置文件的兴趣；及a topic directory comprising a plurality of topics, including topics selectable by a user to indicate interests in one of said topics defining an interest profile of said user; and一列出搜索结果的搜索结果页面，其包括复数个与一用户的一搜索查询相关的文档，并包括一使所述用户能够根据所述用户的所述兴趣配置文件来调整所述文档的一排列的图形控制元件。a search results page listing search results, including a plurality of documents relevant to a user's search query, and including a feature that enables the user to tailor the documents based on the user's interest profile Arrangement of graphic control elements.31、如权利要求30所述的用户界面，其中所述主题目录进一步包括：31. The user interface of claim 30, wherein the subject directory further comprises:每一与复数个网站相关联的主题，其中每一网站均具有一相关联的提升值以用于在所述文档与所述搜索查询相关时提升一与所述网站有关的文档的一信息检索得分。each topic associated with a plurality of websites, each website having an associated boost value for boosting an information retrieval of a document related to the website when the document is relevant to the search query Score.32、如权利要求30所述的用户界面，其中所述图形控制器为一滑标。32. The user interface of claim 30, wherein the graphical control is a slider.33、如权利要求30所述的用户界面，其中所述图形控制器为一刻度盘。33. The user interface of claim 30, wherein the graphical control is a dial.34、如权利要求30所述的用户界面，其中所述图形控制器为一无线电按钮。34. The user interface of claim 30, wherein the graphic control is a radio button.35、如权利要求30所述的用户界面，其中所述图形控制器具有复数个位置，每一位置均与所述用户的兴趣配置文件的一影响程度相关联。35. The user interface of claim 30, wherein the graphical control has a plurality of positions, each position being associated with a degree of influence of the user's interest profile.36、一种搜索引擎系统，其包括：36. A search engine system comprising:一包括复数个主题的主题目录，其包括可由一用户选择的主题以指示一所述主题中用于界定所述用户的一兴趣配置文件的兴趣；a topic directory comprising a plurality of topics, including topics selectable by a user to indicate interests in one of said topics defining an interest profile of said user;一搜索引擎，其用于响应于一由一用户提供的查询来检索包括若干文档的搜索结果并根据与所述查询的相关性来对所述文档进行排列；及a search engine for retrieving search results comprising a number of documents in response to a query provided by a user and ranking the documents according to their relevance to the query; and一个性化组件，其根据所述用户的兴趣配置文件来确定所述文档的至少一种排列。a personalization component that determines at least one permutation of said documents based on said user's interest profile.

Images