Hiep Phuc Luong

Concept-Based Document Recommendations for CiteSeer Authors

The information explosion in todays electronic world has created the need for information filtering techniques that help users filter out extraneous content to identify the right information they need to make important decisions. Recommender systems are one approach to this problem, based on presenting potential items of interest to a user rather than requiring the user to go looking for them. In this paper, we propose a recommender system that recommends research papers of potential interest to authors known to the CiteSeer database. For each author participating in the study, we create a user profile based on their previously published papers. Based on similarities between the user profile and profiles for documents in the collection, additional papers are recommended to the author. We introduce a novel way of representing the user profiles as trees of concepts and an algorithm for computing the similarity between the user profiles and document profiles using a tree-edit distance measure. Experiments with a group of volunteers show that our concept-based algorithm provides better recommendations than a traditional vector-space model based technique.

Ontology-Based Focused Crawling

Ontology learning has become a major area ofresearch whose goal is to facilitate the construction ofontologies by decreasing the amount of effort requiredto produce an ontology for a new domain. However,there are few studies that attempt to automate theentire ontology learning process from the collection ofdomain-specific literature, to text mining to build newontologies or enrich existing ones. In this paper, wepresent a framework of ontology learning that enablesus to retrieve documents from the Web using focusedcrawling in a biological domain, amphibianmorphology. We use a SVM (Support Vector Machine)classifier to identify domain-specific documents andperform text mining in order to extract usefulinformation for the ontology enrichment process. Thispaper reports on the overall system architecture andour initial experiments on the focused crawler anddocument classification.

Conceptual recommender system for CiteSeerX

Short search engine queries do not provide contextual information, making it difficult for traditional search engines to understand what users are really requesting. One approach to this problem is to use recommender systems that identify user interests through various methods in order to provide information specific to the users needs. However, many current recommender systems use a collaborative model based on a network of users to provide the recommendations, leading to problems in environments where network relationships are sparse or unknown. Content-based recommenders can avoid the sparsity problem but they may be inefficient for large document collections. In this paper, we propose a concept-based recommender system that recommends papers to general users of the CiteSeerx digital library of Computer Science research publications. We also represent a novel way of classifying documents and creating user profiles based on the ACM (Association for Computer Machinery) classification tree. Based on these user profiles which are built using past click histories, relevant papers in the domain are recommended to users. Experiments with a set of users on the CiteSeerX database show that our concept-based method provides accurate recommendations even with limited user profile histories.

Publication venue recommendation using author network's publication history

Selecting a good conference or journal in which to publish a new article is very important to many researchers and scholars. The choice of publication venue is usually based on the authors existing knowledge of publication venues in their research domain or the match of the conference topics with their paper content. They may not be aware of new or other more appropriate conference venues to which their paper could be submitted. A traditional way to recommend a conference to a researcher is by analyzing her paper and comparing it to the topics of different conferences using content-based analysis. However, this approach can make errors due to mismatches caused by ambiguity in text comparisons. In this paper, we present a new approach allowing researchers to automatically find appropriate publication venues for their research paper by exploring authors network of related co-authors and other researchers in the same domain. This work is a part of our social network based recommendation research for research publications venues and interesting hot-topic researches. Experiments with a set of ACM SIG conferences show that our new approach outperforms the content-based approach and provides accurate recommendation. This works also demonstrates the feasibility of our ongoing approach aimed at using social network analysis of researchers and experts in the relevant research domains for a variety of recommendation tasks.

Scientific publication recommendations based on collaborative citation networks

To learn about the state of the art for a research project, researchers must conduct a literature survey by searching for, collecting, and reading related scientific articles. Popular search systems, online digital libraries, and Web of Science (WoS) sources such as IEEE Explorer, ACM, SpringerLink, and Google Scholar typically return results or articles that are similar to keywords in the users query. Some digital libraries also include content-based recommenders that suggest papers similar to one the user likes based on the contents of paper, i.e., the keywords it contains. In this work, we present a recommender module that suggests papers to users based on the seed papers Citation Network. This work takes into account the combination of the co-citation and co-reference factors to improve algorithms effectiveness. We applied and improved the the CCIDF (Common Citation Inverse Document Frequency) algorithm used by the CiteSeer digital library. This improved algorithm, called CCIDF+, was evaluated using data collected from Microsoft Academic Search (MAS). Experimental results show that CCIDF+ outperforms CCIDF.

Ontology Learning Through Focused Crawling and Information Extraction

Ontology learning aims to facilitate the construction of ontologies by decreasing the amount of effort required to produce an ontology for a new domain. However, there are few studies that attempt to automate the entire ontology learning process from the collection of domain-specific literature, to text mining to build new ontologies or enrich existing ones. In this paper, we present a complete framework for ontology learning that enables us to retrieve documents from the Web using focused crawling, and then use a SVM (Support Vector Machine) classifier to identify domain-specific documents and perform text mining in order to extract useful information for the ontology enrichment process. We have carried out several experiments on components of this framework in a biological domain, amphibian morphology. This paper reports on the overall system architecture and our initial experiments on information extraction using text mining techniques to enrich the domain ontology.

Ontology Learning Using Word Net Lexical Expansion and Text Mining

In knowledge management systems, ontologies play an important role as a backbone for providing and accessing knowledge sources. They are largely used in the next generation of the Semantic Web that focuses on supporting a better cooperation between humans and ma‐ chines [2]. Since manual ontology construction is costly, time-consuming, error-prone, and inflexible to change, it is hoped that an automated ontology learning process will result in more effective and more efficient ontology construction and also be able to create ontologies that better match a specific application [20]. Ontology learning has recently become a major focus for research whose goal is to facilitate the construction of ontologies by decreasing the amount of effort required to produce an ontology for a new domain. However, most current approaches deal with narrowly-defined specific tasks or a single part of the ontology learn‐ ing process rather than providing complete support to users. There are few studies that at‐ tempt to automate the entire ontology learning process from the collection of domainspecific literature and filtering out documents irrelevant to the domain, to text mining to build new ontologies or enrich existing ones.

Integrating bibliographical data of computer science publications from online digital libraries

In this paper we proposed and developed a system to integrate the bibliographical data of publications in the computer science domain from various online sources into a unified database based on the focused crawling approach. In order to build this system, there are two phases to carry on. The first phase deals with importing bibliographic data from DBLP (Digital Bibliography and Library Project) into our database. The second phase the system will automatically crawl new publications from online digital libraries such as Microsoft Academic Search, ACM, IEEEXplore, CiteSeer and extract bibliographical information (one kind of publication metadata) to update, enrich the existing database, which have been built at the first phase. This system serves effectively in services relating to academic activities such as searching literatures, ranking publications, ranking experts, ranking conferences or journals, reviewing articles, identifying the research trends, mining the linking of articles, stating of the art for a specified research domain, and other related works base on these bibliographical data.

Enriching concept descriptions in an amphibian ontology with vocabulary extracted from wordnet

An important task of ontology learning is to enrich the vocabulary for domain ontologies using different sources of information. WordNet, an online lexical database covering many domains, has been widely used as a source from which to mine new vocabulary for ontology enrichment. However, since each word submitted to WordNet may have several different meanings (senses), existing approaches still face the problem of semantic disambiguation in order to select the correct sense for the new vocabulary to be added. In this paper, we present a similarity computation method that allows us to efficiently select the correct WordNet sense for a concept-word in a given ontology. Once the correct sense is identified, we can then enrich the concepts vocabularly using nearby words in WordNet. Experimental results using an amphibian ontology show that the similarity computation method reach a good average accuracy and our approach is able to enrich the vocabulary of each concept with words mined from WordNet synonyms and hypernyms.

Social network analysis of virtual worlds

As 3D environments become both more prevalent and more fragmented, studying how users are connected via their avatars and how they benefit from the virtual world community has become a significant area of research. An in depth analysis of the virtual world social networks is necessary to evaluate its worlds, to understand the impact of avatar social networks on the virtual worlds, and to improve future online social networks. Our current efforts are focused on building and exploring the social network aspects of virtual worlds. In this paper we evaluate the Second Life social network we have created and compare it to other social networking sites found on the web. Experimental results with data crawled from Second Life virtual worlds demonstrate that our approach was able to build a representative network of avatars in virtual world from the sample data. The analysis comparison between virtual world social networks and others in flat web allows us to gauge measures that better explore the relationship between locations linked by multiple users and their avatars. Using this comparison, we can also determine if techniques of personalization search and content recommendation are feasible for virtual world environments.