Abstract:
Research shows that recommendations comprise a valuable service for users of a digital library [11]. While most existing recommender systems rely either on a content-based approach or a collaborative approach to make recommendations, there is potential to improve recommendation quality by using a combination of both approaches (a hybrid approaches). In this paper, we report how we tested the idea of using a graph-based recommender system that naturally combines the content-based and collaborative approaches. Due to the similarity between our problem and a concept retrieval task, a Hopfield net algorithm was used to exploit high-degree book-book, user-user and book-user associations. Sample hold-out testing and preliminary subject testing were conducted to evaluate the system, by which it was found that the system gained improvement with respect to both precision and recall by combining content-based and collaborative approaches. However, no significant improvement was observed by exploiting high-degree associations.
Abstract:
The advent and rapid proliferation of internet communication has allowed the realization of numerous security issues. The anonymous nature of online mediums such as email, web sites, and forums provides an attractive communication method for criminal activity. Increased globalization and the boundless nature of the internet have further amplified these concerns due to the addition of a multilingual dimension. The world's social and political climate has caused Arabic to draw a great deal of attention. In this study we apply authorship identification techniques to Arabic web forum messages. Our research uses lexical, syntactic, structural, and content-specific writing style features for authorship identification. We address some of the problematic characteristics of Arabic in route to the development of an Arabic language model that provides a respectable level of classification accuracy for authorship discrimination. We also run experiments to evaluate the effectiveness of different feature types and classification techniques on our dataset. © Springer-Verlag Berlin Heidelberg 2005.
Abstract:
Improvised explosive device web pages represent a significant source of knowledge for security organizations. These web pages exist in distinctive genres of communication, providing different types and levels of information for the intelligence community. This paper presents a framework for the classification of improvised explosive device web pages by genre. The approach uses a complex feature extractor, extended feature representation, and support vector machine learning algorithms. Improvised explosive device web pages were collected from the Dark Web and two classification models were examined, one using feature selection. Classification accuracy exceeded 88%. ©2008 IEEE.
Abstract:
As the number of non-English resources available on the Web is increasing rapidly, developing information retrieval techniques for non-English languages is becoming an urgent and challenging issue. In this research to facilitate information seeking in a multilingual world, we focused on discovering how search-engine techniques developed for English could be generalized for use with other languages. We proposed a general framework incorporating a focused collection-building technique, a generic language processing ability, an integration of information resources, and a post-retrieval analysis module. Based on this approach, we developed CMedPort, a Chinese Web portal in the medical domain that not only allows users to search for Web pages from local collections and meta-search engines but also provides encoding conversion between simplified and traditional Chinese to support cross-regional search and document summarization and categorization. User studies were conducted to compare the effectiveness and efficiency of CMedPort with those of three major Chinese search engines. Results indicate that CMedPort achieved similar accuracy for search tasks, but exhibited significantly higher recall than each of the three search engines as well as higher precision than two of the search engines for browse tasks. There were no significant differences among the efficiency measures for CMedPort and benchmarks systems. A post-questionnaire regarding system usability indicated that CMedPort achieved significantly higher user satisfaction than any of the three benchmark systems. The subjects especially liked CMedPort's categorizer, commenting that it helped improve understanding of search results. These encouraging outcomes suggest a promising future for applying our approach to Internet searching and browsing in a multilingual world. © 2005 Elsevier B.V. All rights reserved.
Abstract:
Research has shown that most users' online information searches are suboptimal. Query optimization based on a relevance feedback or genetic algorithm using dynamic query contexts can help casual users search the Internet. These algorithms can draw on implicit user feedback based on the surrounding links and text in a search engine result set to expand user queries with a variable number of keywords in two manners. Positive expansion adds terms to a user's keywords with a Boolean "and," negative expansion adds terms to the user's keywords with a Boolean "not." Each algorithm was examined for three user groups, high, middle, and low achievers, who were classified according to their overall performance. The interactions of users with different levels of expertise with different expansion types or algorithms were evaluated. The genetic algorithm with negative expansion tripled recall and doubled precision for low achievers, but high achievers displayed an opposed trend and seemed to be hindered in this condition. The effect of other conditions was less substantial.
