توجه: محتویات این صفحه به صورت خودکار پردازش شده و مقاله‌های نویسندگانی با تشابه اسمی، همگی در بخش یکسان نمایش داده می‌شوند.
۱Using Q–Learning for Web Recommendations from Web Usage Data
نویسنده(ها): ، ،
اطلاعات انتشار: دوازدهمین کنفرانس سالانه انجمن کامپیوتر ایران، سال
تعداد صفحات: ۴
With the rapid growth of the World Wide Web, the amount of information available online is increasing with an enormous pace. Recommender systems aim at pruning this information space and directing users toward the items that best meet their needs and interests. In this paper we propose a novel machine learning perspective toward the problem, based on reinforcement learning. We model the problem as Q–Learning ,employing concepts and techniques commonly applied in the web usage mining domain. We propose that reinforcement learning paradigm provides an appropriate model for the recommendation problem, as well as a framework in which the system constantly interacts with the user and learns from her behavior. Our experimental evaluations supports our claims and demonstrate how this approach can improve the quality of web recommendations.<\div>

۲A Hybrid Recommender System Applicable for Asynchronous Discussion Groups
نویسنده(ها): ،
اطلاعات انتشار: چهارمین کنفرانس ملی و اولین کنفرانس بین المللی آموزش الکترونیک، سال
تعداد صفحات: ۱۱
Recommender systems have been developed in different domains. One of the important one is information retrieval. In this domain, various systems and architectures have been proposed, that some of them are only based on collaborative filtering and some others benefits from content–based filtering techniques. In other words, these systems either consider the recommendable item information or user’s profile information. But hybrid recommender systems have been considered not as much of other techniques. In hybrid recommender systems the performance is improved by combining existing techniques, and consequently overcoming the weakness of each of the mentioned techniques. In this paper a new hybrid recommender system for asynchronous discussion groups is presented.
Since asynchronous discussion groups play an important role in supporting the collaborative learning methodology, extracting knowledge from discussions is a critical issue in e–learning. Recommender systems could help learners to achieve new knowledge by means of customized information delivered to them. The proposed system is used to recommend useful information from a huge amount of transcripts in group. Considering different recommendation parameters such as user’s past liking, information associated to the content of transcripts, the proposed system is compared with some of the recommender systems in the view of usefulness of suggested recommendations. The main difference between the proposed system and others is in utilizing important parameters related to discussion groups such as "application domain of transcript information". This parameter is useful for any learner who is looking for special information in transcripts such as a programming code, a mathematical formula.<\div>

۳Expert Finding on Social Network with Link Analysis Approach
نویسنده(ها): ، ،
اطلاعات انتشار: نوزدهمین کنفرانس مهندسی برق ایران، سال
تعداد صفحات: ۶
With the appearance of social networks in the Internet, the communications between people took a new form. Nowadays, lots of people with different goals are registered in social networks and do wide range of activities. One of the most important feature of social networks is knowledge sharing. The main problem regarding to this issue is a wide range of shared knowledge and there is no mechanism to determine their validity. So, the knowledge shared on social networks could not be trusted. By finding experts in social networks and determining their level of knowledge, the validity of their posts could be determined. Therefore a solution to the mentioned problem is to provide a method for expert finding. In this research a novel model based on social network analysis is proposed to find the experts who are the members of social networks by means of business intelligence approach. This model is verified by real data from Friendfeed social network. First, data is extracted, transformed and loaded to data warehouse with ETL processes. Then a new ranking algorithm is proposed for finding the experts, and finally the obtained results are compared to the experts' opinions utilizing spearman's correlation function.<\div>

۴TagAssessment: Using Tagging Technology for Learner Assessment
اطلاعات انتشار: ششمین کنفرانس ملی و سومین کنفرانس بین المللی یادگیری و آموزش الکترونیک، سال
تعداد صفحات: ۶
Nowadays, with development of Information technology, using e–learning for higher quality of education has been increased. On the other hand with the emergence of Web 2.0 and high interest for using its technologies such as tagging, annotating and social networking caused to more users participation than before. In this regard, one of the most popular technologies is Tagging. Usually users used to place tags on the contents as a reminder in future. In this paper, a novel method is proposed to assess the learners’ understandings. The mentioned method which is named TagAssessment computes the figure of relationship between the contents and the user’s tags and utilizes the obtained results for learner assessment. In this regard, the semantic distance between the tags associated to the content, and the concepts covered by the same content is calculated using WordNet tree ontology. In the evaluation process, ninety learners were asked to place their tags on the contents related to a BSc course to create a data set for further evaluation. The obtained results showed that as the smaller the distance between the learners’ tags and the contents are the better comprehension have been achieved.<\div>

۵XMLCMAP: A Tool for Combining Concept Maps for Automatic Knowledge Extraction
نویسنده(ها): ،
اطلاعات انتشار: ششمین کنفرانس ملی و سومین کنفرانس بین المللی یادگیری و آموزش الکترونیک، سال
تعداد صفحات: ۶
Nowadays there are various ways to represent knowledge among which the concept mapping technique is the one of the most practical of all and the usage of concept mapping to represent a subject keeps increasing day by day. The important problem is the lack of an adequate tool to automatically extract the present knowledge from concept maps and also the absence of a tool to combine these maps in order to find new knowledge or to produce comprehensive maps. The current paper introduces a tool called XMLCMAP which has four essential abilities: saving and retrieving the concept maps with XML format; Automatic extraction of existing knowledge from a concept map; the combination of the maps to produce a comprehensive map; and getting new knowledge from previously combined maps. The function of this tool is based on converting maps to XML structures, processing these XML structures, and doing some queries. This tool which has an important role in knowledge extraction is able to make comprehensive maps quite quickly according to the needs of the users. At the end of this paper, the degree of tool performance is assessed and the results and future works are cited.<\div>

۶A Framework for Learner Preparation and Support in Academic E–learning Environment
اطلاعات انتشار: International Journal Information and Communication Technology Research، سوم،شماره۱، Jan ۲۰۱۱، سال
تعداد صفحات: ۹
E–learning as a modern educational system has certain components, and adopting this system will affect the conditions and characteristics of administrative aspects of the educational systems. Some of the critical educational factors such as requirements\ necessary skills \capabilities\ admission factors and learner capabilities \ characteristics will evolve significantly in this modern educational paradigm.The main objective of this paper is to present a framework for main indices and factors regarding e–learner readiness evaluation and also identifying the learners’ skills, capabilities and limitations in a way to enable e–learning system administrators to provide preparation and support services for improving learning quality. In this paper, through a survey of e–learners’ readiness characteristics, a framework for e–learners’ primary preparation and corresponding guidelines for e–learners’ support is proposed which is based on level of readiness. The proposed framework is targeted to academic courses, though its analysis mechanism can be easily workable for other educational courses as well.

۷Expert Finding System in E–Learning: Using Concept Maps (Research Note)
نویسنده(ها): ،
اطلاعات انتشار: International Journal Information and Communication Technology Research، سوم،شماره۱، Jan ۲۰۱۱، سال
تعداد صفحات: ۹
Finding subject experts for problem solving is an important issue in e–learning environment. In e–learningenvironment there is no direct way to find the superior individuals. The current methods like analyzing the discussions or considering the learner actions need a lot of data or have some limitations. In this work, concept maps are utilized to define the experts in an e–learning environment. Concepts maps are graphical explanation of the meaningful relationship among the concepts. 35 M.Sc. and B.Sc. students participated in this work. The test design, implemented software, results and conclusions are described in detail in this article. The novelty of this work is the application of concept maps to find the experts, the combined presented models for concept map construction, and presenting a research methodology in this area.
نمایش نتایج ۱ تا ۷ از میان ۷ نتیجه