Daifeng Li

Community-based topic modeling for social tagging

Exploring community is fundamental for uncovering the connections between structure and function of complex networks and for practical applications in many disciplines such as biology and sociology. In this paper, we propose a TTR-LDA-Community model which combines the Latent Dirichlet Allocation model (LDA) and the Girvan-Newman community detection algorithm with an inference mechanism. The model is then applied to data from Delicious, a popular social tagging system, over the time period of 2005-2008. Our results show that 1) users in the same community tend to be interested in similar set of topics in all time periods; and 2) topics may divide into several sub-topics and scatter into different communities over time. We evaluate the effectiveness of our model and show that the TTR-LDA-Community model is meaningful for understanding communities and outperforms TTR-LDA and LDA models in tag prediction.

Adding community and dynamic to topic models

The detection of communities in large social networks is receiving increasing attention in a variety of research areas. Most existing community detection approaches focus on the topology of social connections (e.g., coauthor, citation, and social conversation) without considering their topic and dynamic features. In this paper, we propose two models to detect communities by considering both topic and dynamic features. First, the Community Topic Model (CTM) can identify communities sharing similar topics. Second, the Dynamic CTM (DCTM) can capture the dynamic features of communities and topics based on the Bernoulli distribution that leverages the temporal continuity between consecutive timestamps. Both models were tested on two datasets: ArnetMiner and Twitter. Experiments show that communities with similar topics can be detected and the co-evolution of communities and topics can be observed by these two models, which allow us to better understand the dynamic features of social networks and make improved personalized recommendations.

LIS Dissertation Titles and Abstracts (1930-2009): Where Have All the Librar* Gone?

This article examines the topicality of Library and Information Science (LIS) dissertations written between 1930 and 2009 at schools with American Library Association (ALA)–accredited university programs in North America. Dissertation titles and abstracts were examined for the presence of library-related keywords drawn from the core curricula of ALA-accredited schools, and trend data were created to describe the evolution of LIS doctoral research over the past eighty years. The results show that the percentage of dissertations found to contain no instance of any of the selected library keywords has steadily risen since 1980. Simultaneously, the percentage of dissertations found to contain instances of keywords in both the title and abstract has steadily declined. The results provide general empirical support for long-held anecdotal assertions that libraries are no longer the primary research focus at the doctoral level in LIS.

Semantic web portal: a platform for better browsing and visualizing semantic data

One of the main shortcomings of Semantic Web technologies is that there are few user-friendly ways for displaying, browsing and querying semantic data. In fact, the lack of effective interfaces for end users significantly hinders further adoption of the Semantic Web. In this paper, we propose the Semantic Web Portal (SWP) as a light-weight platform that unifies off-the-shelf Semantic Web tools helping domain users organize, browse and visualize relevant semantic data in a meaningful manner. The proposed SWP has been demonstrated, tested and evaluated in several different use cases, such as a middle-sized research group portal, a government dataset catalog portal, a patient health center portal and a Linked Open Data portal for bio-chemical data. SWP can be easily deployed into any middle-sized domain and is also useful to display and visualize Linked Open Data bubbles.

Dynamic Features of Social Tagging Vocabulary: Delicious, Flickr and YouTube

This article investigates the dynamic features of social tagging vocabularies in Delicious, Flickr and YouTube from 2003 to 2008. It analyzes the evolution of the usage of the most popular tags in each of these three social networks. We find that for different tagging systems, the dynamic features reflect different cognitive processes. At the macro level, the tag growth obeys power-law distribution for all three tagging systems with exponents lower than one. At the micro level, the tag growth of popular resources in all three tagging systems follows a similar power-law distribution. Moreover, we find that the exponents of tag growth varied in different evolving stages of popular individual resources.