Adapting domain ontology for personalized knowledge search and recommendation

This paper presents research on the development of a domain ontology adaptation system for personalized knowledge search and recommendation that adapts a suitable domain ontology according to the previous browsing and reading behavior of users (i.e., usage history log). An adaptive domain ontology can satisfy the future requirements of users and promote use value. In developing the system, a domain ontology adaptation model is first designed. Based on the designed adaptation model, a methodology for domain ontology adaptation is developed. Subsequently, a domain ontology adaptation system is implemented with an illustrative example of securities trading. Finally, a system evaluation for user satisfaction and a methodology evaluation are conducted to demonstrate that the developed methodology and system worked efficiently.     

A personalized recommendation procedure for Internet shopping support

The rapid growth of e-commerce has caused product overload where the customer is no longer able to effectively choose the products he/she is exposed to. To overcome the product overload of Internet shoppers, several recommender systems have been developed. Recommendation systems track past actions of a group of customers to make a recommendation to individual members of the group. We introduce a personalized recommendation procedure by which we can get further recommendation effectiveness when applied to Internet shopping malls. The suggested procedure is based on Web usage mining, product taxonomy, association rule mining, and decision tree induction. We applied the procedure to a leading Internet shopping mall in Korea for performance evaluation, and some experimental results are provided. The experimental results show that choosing the right level of product taxonomy and the right customers increases the quality of recommendations.     

Assessing users'' product-specific knowledge for personalization in electronic commerce

While many electronic commerce (EC) companies are adopting one-to-one marketing approaches using various personalization technologies to make their products and services unique for the purpose of attracting and retaining customers and improving their completion edges in the EC ecosystem, which, nevertheless, has low entrance barriers for new players to join and further intensify the competition, none or few of them consider a fundamental issue—the user's product-specific knowledge. Our research proposed to add this new domain of the customer's knowledge on appropriate target products into the personalization process as a part of the overall EC strategy for businesses. In this paper, we present our initial design for assessing the user's product-specific knowledge using the proposed innovative method for detecting it directly in a non-intrusive way without asking users to answer or fill out any types of questionnaires. Our method is based on customer's on-line navigation behaviors by analyzing their navigation patterns through pre-trained artificial neural networks. An empirical study designed for a case of EC store selling digital cameras was conducted in our research to prove the concept, and a good preliminary result was derived from the study.

For the purpose of comparing the performances between the conventional approach of using questionnaire and the proposed innovative approach of navigation pattern mining, a questionnaire based approach for evaluating the user's product-specific knowledge was designed and incorporated into our knowledge level assessment system (KLAS). Our study result shows that although the pure questionnaire-based KLAS is intrusive and may not be accepted by some users, for those users willing to complete the questionnaire, the proposed navigation pattern approach can be combined with the questionnaire-based approach to create a hybrid KLAS which has a significantly improved accuracy rate in detecting the customer's product knowledge level.     

基于领域知识的个性化协同商务推荐系统

基于领域知识与顾客购买倾向相关联的事实,从知识表示、知识获取、系统实现三个方面研究了个性化协同商务推荐系统的实现策略。知识表示研究了自然语言的本体表示,主要包括：知识本体描述、模糊关系设计、概念关联抽象和公理修正四个部分;知识获取采用多层次领域知识获取和基于数据挖掘的智能知识获取两种方法,对知识的形式化和结构化进行了研究;基于J2EE技术创建了由客户端、服务器端、存储系统组成的协同商务推荐系统的结构模型。最后通过测试网站对系统的有效性进行了验证。     

融合知识图谱和评论文本的个性化推荐模型

推荐系统的目的是解决“信息过载”的问题,然而目前的研究方法大多利用用户和商品信息对用户兴趣进行建模,没有同时利用知识图谱构建用户模型来增强推荐系统的性能,因此提出了融合知识图谱和评论文本的个性化推荐模型。首先,通过不同类型的知识图谱分别关联用户项目和用户评论文本,扩展用户的兴趣和提取评论文本中的实体;其次,通过构造用户兴趣网络得到带有用户兴趣偏好的兴趣特征;然后,通过构造画像模块和情感模块的画像网络提取到带有用户情感倾向的画像特征;利用决策层进行点击率预测。最后在Amazon数据集上进行了实验比较,对所提模型的性能进行了评估,并与目前的融合知识图谱和评论文本的推荐模型进行比较,验证了所提模型的有效性。     

面向个性化推荐的node2vec-side融合知识表示

推荐系统中知识图谱对系统的推荐效果起到很重要的作用,图谱中的知识表示成为影响推荐系统的关键因素,这也成为当前的研究热点之一。针对推荐系统中知识图谱的结构特点,在传统node2vec模型基础上增加关系表示和多元化游走策略,提出一种基于node2vec的知识表示node2vec-side,结合推荐系统知识图谱网络结构,旨在挖掘大规模推荐实体节点间潜在的关联关系,降低表示方式复杂度,提高可解释性。经过时间复杂度分析可知,提出的知识表示方式在复杂度上低于Trans系列和RGCN。在传统知识图谱数据集FB15K、WN18和推荐领域数据集MovieLens-1M、Book-Crossing、Last.FM上分别进行链接预测对比实验。实验结果表明,在MovieLens-1M数据集上,Hits@10分别提升了5.5%～12.1%,MRR提升了0.09～0.24;在Book-Crossing数据集上,Hits@10分别提升了3.5%～20.6%,MRR平均提升了0.04～0.24;而在Last.FM数据集上,hits@1提升了0.3%～8.5%,MRR平均提升了0.04～0.16,优于现有算法,验证了所提方法的有效性。     

A real-time personalized route recommendation system for self-drive tourists based on vehicle to vehicle communication

Recently, traffic jams and long queuing problems in tourist hot spots is growing with the increasing number of self-drive tourists. Some recommendation systems have been developed in attempt to relieve these problems. However, all these systems lack information pertaining to real-time traffic as well as the ability of personalization. In this research, we have developed a novel route recommendation system to provide self-drive tourists with real-time personalized route recommendations. This will help to reduce the traffic jams and queuing time in tourist hot spots. It will also help to personalize visiting routes based on the user’s specific preferences. Ultimately, based on the evaluation results given by experienced self-drive tourists, we have shown that the proposed system not only saves total visiting time, but also meets their specific visiting preferences.     

基于CoP建模的协作过滤推荐方法

传统的协作过滤推荐方法主要基于个人兴趣特征来实现推荐。在组织内部协作场景下,为实现知识共享与重用,推荐系统不仅要考虑用户兴趣,还应考虑用户和用户组的任务。传统的协作过滤推荐方法已不能满足要求。CoP是组织内部人员管理的主要形式,它的特征是其成员任务的反映。基于已有的协作过滤推荐研究与D-S理论,提出了一种CoP特征构建算法,并以此为基础研究了面向CoP的协作过滤推荐。     

A framework for applying intelligent agents to support electronic trading

The purpose of this paper is to study how Intelligent Agents (IAs) can be used to facilitate electronic trading. An IA is a software program designed for performing a specific task based on its own knowledge and the message it received. Given the increased complexity of Internet services, many IAs are useful to make electronic markets more effective.In the paper, activities and structures of electronic markets are reviewed and discussed with respect to the coordination mechanism and primitive activities. This is followed by an analysis of IAs useful for electronic commerce (EC). A three-layer architecture for organizing IAs for EC is developed. Finally, application of the framework to support EC and related issues are presented. The findings are useful for implementing a more effective environment for EC.     

一种面向学习路径推荐的知识网络构建方法

随着在线教育的迅速发展,互联网上的教学资源数量也呈现出快速增长的趋势。针对当前在线学习平台普遍存在着教学资源内容重复、人们难以辨别与选择,导致学习者很难应用这些资源构建适合自己的学习路径的问题,提出一种面向学习路径推荐的领域知识网络构建方法。通过对每个学习对象的预备知识与目标知识进行社会标注,构建相应的领域知识网络,然后,运用弗洛伊德算法计算领域知识网络里任意两个知识点间的最短路径,为学习路径推荐提供基础。     

Web服务个性化推荐研究综述

随着Web服务的广泛使用和互联网上服务数量的增加,如何向用户提供最佳的服务选择列表成为了新的挑战.Web服务个性化推荐实现了由被动接受用户请求向主动感知用户需求的转变.个性化的Web服务推荐方法已经成为Web服务发现和选择的有效辅助手段.Web服务的个性化推荐技术也成为了近年来服务计算领域的研究热点.对当前Web服务个性化推荐的文献进行了归类分析,总结了当前Web服务个性化推荐的技术现状、研究方法和实验的数据集,列出了未来Web服务个性化推荐研究热点和挑战.     

Modeling and broadening temporal user interest in personalized news recommendation

User profiling is an important step for solving the problem of personalized news recommendation. Traditional user profiling techniques often construct profiles of users based on static historical data accessed by users. However, due to the frequent updating of news repository, it is possible that a user’s fine-grained reading preference would evolve over time while his/her long-term interest remains stable. Therefore, it is imperative to reason on such preference evaluation for user profiling in news recommenders. Besides, in content-based news recommenders, a user’s preference tends to be stable due to the mechanism of selecting similar content-wise news articles with respect to the user’s profile. To activate users’ reading motivations, a successful recommender needs to introduce “somewhat novel” articles to users.In this paper, we initially provide an experimental study on the evolution of user interests in real-world news recommender systems, and then propose a novel recommendation approach, in which the long-term and short-term reading preferences of users are seamlessly integrated when recommending news items. Given a hierarchy of newly-published news articles, news groups that a user might prefer are differentiated using the long-term profile, and then in each selected news group, a list of news items are chosen as the recommended candidates based on the short-term user profile. We further propose to select news items from the user–item affinity graph using absorbing random walk model to increase the diversity of the recommended news list. Extensive empirical experiments on a collection of news data obtained from various popular news websites demonstrate the effectiveness of our method.     

基于情境相似的知识个性化推荐系统研究

知识情境是知识创造和运用的具体环境和背景,融合知识情境的知识个性化推荐系统是提高知识重用效率和共享特性的重要手段。提出了在知识个性化推荐系统中添加知识情境,使用多层多维度建模方法构建知识情境模型,通过知识情境的相似性评估,将与当前目标情境相似度满足特定值的历史情境所关联的知识推荐给目标用户。实验表明,此方法一定程度上能提高知识个性化推荐的效率。     

面对智能导诊的个性化推荐算法

针对电子医疗信息过载和医疗资源严重不足的问题,本文以辅助诊疗的结果为基础,将Skyline查询和局部范围内基于协同过滤的评分方式有机结合,提出了一种面向智能导诊的个性化推荐算法。实验结果表明,本文提出的算法能为用户提供个性化的合理推荐结果。该方法对合理分配和使用医疗资源有很大的促进作用,能从一定程度上缓解就诊压力,提高就诊质量,具有重要的实用价值和社会意义。     

A collaborative filtering-based approach to personalized document clustering

Document clustering is an intentional act that reflects individual preferences with regard to the semantic coherency and relevant categorization of documents. Hence, effective document clustering must consider individual preferences and needs to support personalization in document categorization. Most existing document-clustering techniques, generally anchoring in pure content-based analysis, generate a single set of clusters for all individuals without tailoring to individuals' preferences and thus are unable to support personalization. The partial-clustering-based personalized document-clustering approach, incorporating a target individual's partial clustering into the document-clustering process, has been proposed to facilitate personalized document clustering. However, given a collection of documents to be clustered, the individual might have categorized only a small subset of the collection into his or her personal folders. In this case, the small partial clustering would degrade the effectiveness of the existing personalized document-clustering approach for this particular individual. In response, we extend this approach and propose the collaborative-filtering-based personalized document-clustering (CFC) technique that expands the size of an individual's partial clustering by considering those of other users with similar categorization preferences. Our empirical evaluation results suggest that when given a small-sized partial clustering established by an individual, the proposed CFC technique generally achieves better clustering effectiveness for the individual than does the partial-clustering-based personalized document-clustering technique.     

基于改进规则引擎的农业知识推荐系统

为了使农民更加高效的获取农业知识,提出了一种基于改进规则引擎的农业知识系统.通过深入研究规则引擎的工作原理,将其引入到农业知识推荐系统中,增加了系统的正确性和准确性,使系统能够在最恰当的时间向农民提供他们最希望得到的正确的农业知识.同时根据农业领域的特点对规则引擎进行了改进,提出了规则库的树形结构化,并采用规则文件的运行前编译方式.将改进后的规则引擎应用于农业知识推荐系统中,提高了系统的效率.     

基于知识图嵌入的协同过滤推荐算法

利用知识图谱进行推荐的一个巨大挑战在于如何获取项目的结构化知识并对其进行语义特征提取.针对这一问题,提出了一种基于知识图嵌入的协同过滤推荐算法(KGECF).首先从Freebase知识图谱中提取与项目相关的知识信息,并与历史交互项目进行链接构建子知识库;然后通过基于TransR的Xavier-TransR方法得到子知识库中实体、关系表征;设计一种端到端的联合学习模型,将结构化信息与历史偏好信息嵌入到统一的向量空间中;最后利用协同过滤方法进一步计算这些向量并生成精确的推荐列表.在MovieLens-1 M和Amazon-book两个公开数据集上的实验表明,该算法在推荐准确率、召回率、F1值和NDCG四个指标上均优于基线方法,能够集成大规模的结构化和非结构化数据,同时获得高精度的推荐结果.     

基于知识的电子商务智能推荐系统平台设计

分析了传统推荐技术存在的不足,阐述了基于知识的推荐技术的特点及其发展。针对现有基于知识的电子商务推荐系统中存在的不足,提出了基于知识的电子商务智能推荐需要解决的基本问题,设计了基于知识的电子商务智能推荐平台的逻辑框架,并阐述了其工作原理。     

Towards a journalist-based news recommendation system: The Wesomender approach

The present paper introduces a context-aware recommendation system for journalists to enable the identification of similar topics across different sources. More specifically a journalist-based recommendation system that can be automatically configured is presented to exploit news according to expert preferences. News contextual features are also taken into account due to the their special nature: time, current user interests, location or existing trends are combined with traditional recommendation techniques to provide an adaptive framework that deals with heterogeneous data providing an enhanced collaborative filtering system. Since the Wesomender approach is able to generate context-aware recommendations in the journalism field, a quantitative evaluation with the aim of comparing Wesomender results with the expectations of a team of experts is also performed to show that a context-aware adaptive recommendation engine can fulfil the needs of journalists daily work when retrieving timely and primary information is required.