返回项目页面
对比文件列表
2000-04-25_US6055569A_发明授权_US6055569A Accelerating web access by predicting user action_+++A_B_E_F_J_M_c+++.docx
2026-06-17 01:37
2002-09-05_US2002124075A_发明申请_US20020124075A1 Probability associative matrix algorithm_+++A_B_E_F_G_j_m+++.docx
2026-06-17 01:37
2002-09-19_CA2437933A_发明专利_CA2437933A1 PREDICTIVE CACHING AND HIGHLIGHTING OF WEB PAGES_+++a_b_c_d_e_f_g_h_j_l+++.docx
2026-06-17 01:37
2002-11-07_US2002165925A_发明申请_US20020165925A1 Predictive caching and highlighting of web pages_+++A_B_C_H_J_e_l+++.docx
2026-06-17 01:37
2003-05-27_US6571279B_发明授权_US06571279B1 Location enhanced information delivery system_+++A_B_C_E_F_G_J_K_M+++.docx
2026-06-17 01:37
2004-05-27_US2004103371A_发明申请_US20040103371A1 Small form factor web browsing_+++A_B_E_F_G_H_I_J_K_L_c_d_m+++.docx
2026-06-17 01:37
2005-03-31_US2005071864A_发明申请_US20050071864A1 Systems and methods for using interaction information to deform representations of digital content_+++A_B_E_h_l+++.docx
2026-06-17 01:37
2005-08-11_US2005177414A_发明申请_US20050177414A1 Method and apparatus for automatically and continuously pruning prediction models in real time based on data mining_+++C_E_F_M+++.docx
2026-06-17 01:37
2006-02-02_WO2006012144A_发明申请_WO2006012144A2 ACCELERATING USER INTERFACES BY PREDICTING USER ACTIONS_+++A_B_C_D_E_F_G_J_K+++.docx
2026-06-17 01:37
2006-02-23_US2006041548A_发明申请_US20060041548A1 System and method for estimating user ratings from user behavior and providing recommendations_+++A_C_E_F_b+++.docx
2026-06-17 01:37
2007-04-10_US7203909B_发明授权_US07203909B1 System and methods for constructing personalized context-sensitive portal pages or views by analyzing patterns of users' information access activities_+++A_B_C_E_F_G_M+++.docx
2026-06-17 01:37
2007-05-29_US7225397B_发明授权_US07225397B2 Display annotation and layout processing_+++E_H_J_K_L+++.docx
2026-06-17 01:37
2007-12-27_US2007298843A_发明申请_US20070298843A1 Mobile terminal and method for displaying standby screen according to analysis result of user's behavior_+++C_E_F_m+++.docx
2026-06-17 01:37
2008-01-02_CN101098530A_发明公开_CN101098530A 根据用户行为分析结果显示待机屏幕的移动终端及方法_+++C_E_f_m+++.docx
2026-06-17 01:37
2008-03-25_US7349747B_发明授权_US07349747B2 Monitoring device and monitoring method for injection molding machine_+++C_E_F_m+++.docx
2026-06-17 01:37
2008-04-17_US2008092078A_发明申请_US20080092078A1 Scroll Position Estimation Apparatus and Method_+++A_B_C_D_E_F_G_H_I_M+++.docx
2026-06-17 01:37
2008-08-21_US2008201242A_发明申请_US20080201242A1 Dynamic reconfiguration of web pages based on user behavioral portrait_+++A_C_E_F_b_d_j+++.docx
2026-06-17 01:37
2009-06-03_CN101446979A_发明公开_CN101446979A 动态热点跟踪的方法_+++A_B_C_E_M_f_j+++.docx
2026-06-17 01:37
2009-06-25_US2009164395A_发明申请_US20090164395A1 MODELING, DETECTING, AND PREDICTING USER BEHAVIOR WITH HIDDEN MARKOV MODELS_+++A_B_E_F_G_c_k_m+++.docx
2026-06-17 01:37
2006-02-23_US2006041548A_发明申请_US20060041548A1 System and method for estimating user ratings from user behavior and providing recommendations_+++A_C_E_F_b+++.docx

对比文件的名称:2006-02-23_US2006041548A_发明申请_US20060041548A1 System and method for estimating user ratings from user behavior and providing recommendations

目标专利的名称:963浏览建模的方法和装置CN102473187B

本次调用的模型名称:无

### 特征比对表格

技术特征描述以及公开性判断结果对比文件原文引用公开性论述
**技术特征A:其包括:收集与导航行为相对应的数据** <br> **《直接公开》****[0006]** “Automated mechanisms have the potential to help users locate the information and/or products they are seeking.” <br> **[0056]** “In the recommendation system presented herein, ratings for seen items may be inferred from, among other things, the behavior of the user and properties of the seen items.” <br> **[0056]** “...a far from exhaustive list of exemplary behaviors of the user for an online embodiment of the invention may include: the user's viewing times of the seen items, user interaction with multimedia features of pages, mouse movement and backtracking.”对比文件明确公开了一种基于用户行为来估计用户评级的系统。该系统“收集”与用户“行为”相对应的数据,例如“查看时间”、“鼠标移动”和“回溯”。这些行为(例如鼠标移动和回溯)与目标专利中的“导航行为”本质相同,都属于用户在浏览过程中产生的动作。对比文件的目的正是利用这些行为数据来推断用户对物品的喜好,因此明确公开了“收集与导航行为相对应的数据”。
**技术特征B:所述导航行为涉及对浏览器应用的页面进行导航** <br> **《隐含公开》****[0064]** “...the interface 14 is somewhat more distributed and is provided by a computer 14a that connects, via a network connection 14b (such as the internet for example), to a website 14c that is hosted on a web server 14d.” <br> **[0063]** “In use, the interface 14 provides the recommendation system 10 with information on the actions of the user 12 with relation to at least one item in the item data store 16. This information is generally referred to herein as training data...”对比文件虽然未明确使用“浏览器应用”一词,但其系统用于在线环境(online context),用户通过计算机与“网站”交互以实现对项目(item)的浏览。在在线环境中,用户浏览网站必然是通过浏览器应用来实现的。本领域技术人员可以合理推断,当用户通过计算机上的接口(如浏览器)连接到网站并进行导航行为(如查看页面、移动鼠标等)时,其行为必然是“对浏览器应用的页面进行导航”。因此,该特征被隐含公开。
**技术特征C:其中所述数据包括计时数据** <br> **《直接公开》****[0079]** “Some examples of rating predictor variables include viewing time, familiarity, surprisingness, and complexity.” <br> **[0057]** “consider a user interacting with an electronic repository... In general, one would expect users to look at items they like, or find interesting, for a longer time than items that they do not like...Thus, viewing times may contribute to generating implicit ratings...”对比文件公开了“浏览时间(viewing time)”是评级预测变量之一,并且用户的“浏览时间”被用于生成隐式评级。浏览时间是一种典型的计时数据,因此对比文件明确公开了数据包括计时数据。
**技术特征D:所述计时数据与在导航期间何时显示所述页面内的区域相对应** <br> **《未公开》****[0057]** “...users will view relevant information for a longer time (and read more of it) than irrelevant information.” <br> **[0056]** “...a far from exhaustive list of exemplary behaviors of the user for an online embodiment of the invention may include: the user's viewing times of the seen items...”对比文件公开的计时数据(浏览时间)是针对“一个项目(an item)”的整个页面或物品页面的。例如,用户查看一本书的页面、一个商品的页面。对比文件关注的是用户对完整“物品(item)”的查看时间,并未公开或暗示将计时数据与“页面内的区域(例如页面中的某个特定子区域)”进行关联或对应。对比文件中的行为数据(如查看时间)是从整个页面层面采集的,而非页面内的特定区域。因此,该特征未被公开。
**技术特征E:启动对所述数据的存储** <br> **《直接公开》****[0077]** “The input to the ratings estimation module 20 is the seen items and the rating predictor variables (xi's) which may be based on at least one user-specific property. The seen items are provided in the training set data...” <br> **[0079]** “In general, the training set T may be a two-dimensional array in which the ith row contains information about the ith item in the training set T...” <br> **[0089]** “All of the processed information for the text and numeric attributes is stored in the user profile (u).”对比文件公开了收集的行为数据被用作“训练集数据(training set data)”以生成“训练集T”,并最终生成并“存储”在“用户画像(user profile u)”中。例如,用户的查看时间等评级预测变量被存储在训练集数组中。这明确公开了“启动对数据的存储”。
**技术特征F:基于所述数据生成预测模型** <br> **《直接公开》****[0082]** “The pre-processed training set data is then sent to the user ratings estimator 28 which estimates ratings for at least some of the seen items... One method is to use a linear combination, as exemplified in equation 1... where c is a constant, each ai is a coefficient corresponding to the ith rating predictor variable, and each xi is the value of the ith rating predictor variable.” <br> **[0115]** “An example of a preference regression model is shown in Table 8. The preference regression model has an R2 of 0.300...”对比文件公开了使用收集的数据(训练集数据)来“生成模型”。具体而言,其通过线性回归等统计技术建立了一个预测模型(如偏好回归模型),模型中的系数(a_i)是基于数据计算得出的。该模型用于估计用户对物品的评级。因此,这明确公开了“基于数据生成预测模型”。
**技术特征G:执行所述预测模型以便基于所存储的数据来预测所述浏览器应用的另一页面或所述页面内的所述区域。** <br> **《未公开》****[0084]** “The recommendation module 22 assigns an estimated rating to some or all of the items in the item set 18 that the user 12 has not seen and provides a set of recommended items to the user 12.” <br> **[0102]** “The item set rating estimator 34 then generates the list of recommended items Ω for presentation to the user 12.”对比文件公开了执行模型来为“未见过的物品(unseen items)”预测“评级(rating)”,并将评级高的物品作为推荐列表呈现给用户。然而,其预测的对象是“物品(item)”,而非目标专利中的“页面(page)”或“页面内的区域(region within the page)”。目标专利的预测目的是导航辅助(如直接跳转到页面的某个区域),而对比文件的目的是物品推荐。虽然对比文件预测了用户的未来行为(选择哪个物品),但其预测内容和应用场景与目标专利存在本质区别。因此,该特征未被公开。
**技术特征H:其中所述数据进一步包括:页面结构数据,其与所述页面的布局相对应** <br> **《未公开》****全文**对比文件通篇未涉及“页面结构数据”或“页面布局”的概念。对比文件处理的数据是“物品(item)”的“属性(attribute)”,例如书的作者、价格、流派等,以及用户的“行为(behavior)”数据。其数据结构是基于物品属性的,而非浏览器页面渲染后的文档对象模型(DOM)结构。因此,该特征未被公开。
**技术特征I:页面查看区域数据,其将所述页面内的所述区域与所述页面结构数据进行相关。** <br> **《未公开》****全文**对比文件没有公开任何关于“页面查看区域”、“页面结构数据”以及将两者进行“相关”的技术内容。对比文件不关注用户在页面内部的具体查看区域,也不存在将区域与页面结构(如DOM子树)关联的机制。因此,该特征未被公开。
**技术特征J:其进一步包括:接收对所述另一页面的请求** <br> **《未公开》****[0063]** “In use, the interface 14 provides the recommendation system 10 with information on the actions of the user 12 with relation to at least one item in the item data store 16.” <br> **[0078]** “The set of items seen by the user 12 at the time the recommendation process is begun is included in the training set T...”对比文件公开了用户的行为动作作为输入,但并未明确公开“接收对另一页面的请求”。在对比文件的方案中,推荐系统是基于已浏览的物品自动触发预测和推荐,而非在接收到用户访问新页面的请求后才进行预测。用户访问新页面的行为可能被隐式地理解为一种请求,但对比文件未将其作为系统执行预测的触发条件进行描述。
**技术特征K:确定出所述另一页面和所述数据包括在预定门限内类似的页面结构。** <br> **《未公开》****[0075]** “Text attributes in the internal item data store can be standardized or categorized so that the recommendation system 10 knows that these items are the same thing.” <br> **[0092]** “...a similar concept can be used for text or categorical attributes to exclude items with attributes having certain values by applying a threshold to a related numeric attribute.”对比文件存在“相似性”判断,但判断的是“物品”的“属性”相似性,而非“页面”的“页面结构(如DOM结构)”相似性。对比文件通过比较物品的属性值(如文本属性、数值属性)来确定相似性,这与目标专利中通过比较页面结构(如树编辑距离)来确定两个页面结构是否类似的方案完全不同。因此,该特征未被公开。
**技术特征L:其中,所述页面结构数据对应于文档对象模型格式。** <br> **《未公开》****全文**对比文件全文未提及“文档对象模型(DOM)”或任何与HTML/XML文档结构相关的数据格式。其处理的数据是基于关系型数据库或类似结构的物品属性数据,与DOM格式无关。因此,该特征未被公开。
**技术特征M:其中,所述模型定期地或者当收集了预定门限量的所述数据时被更新。** <br> **《未公开》****[0077]** “The rating estimation module 20 may be invoked after the user 12 has viewed a pre-specified number of items from the item data store 16.” <br> **[0077]** “The recommendation system 10 may be fine-tuned by specifying a minimum number of items that need to be viewed by the user 12 before the recommendation system 10 provides a recommended set of items.”对比文件公开了在用户查看了“预设数量(pre-specified number / minimum number)”的物品后,才调用评级估计模块或提供推荐。这表明推荐过程是在收集到一定阈值的数据量后才触发的。然而,这描述的是“使用模型”的触发条件,而非对“模型本身进行更新”的触发机制。对比文件并未公开在模型生成后,会“定期地”或“当收集了新的预定门限量的数据时”对模型进行重新训练或更新。虽然理论上可以更新,但对比文件未予以明确或暗示。

### 结论

综上分析,对比文件直接公开了技术特征 A、C、E、F。

对比文件隐含公开了技术特征 B。

对比文件未公开技术特征 D、G、H、I、J、K、L、M。

<<<A>>><<<C>>><<<E>>><<<F>>><<<b>>>

Powered by Django

网站备案号:渝ICP备2023012882号


重庆市非显而易见网络科技有限责任公司 A Anti NPE NPE