Link recommendation has attracted significant attentions from both industry practitioners and academic researchers. In industry, link recommendation has become a standard and most important feature in online social networks, prominent examples of which include People You May Know on LinkedIn and You May Know on Google+. In academia, link recommendation has been and remains a highly active research area. This paper surveys state-of-the-art link recommendation methods, which can be broadly categorized into learning-based methods and proximity-based methods. We further identify social and economic theories, such as social interaction theory, that underlie these methods and explain from a theoretical perspective why a link recommendation method works. Finally, we propose to extend link recommendation research in several directions that include utility-based link recommendation, diversity of link recommendation, link recommendation from incomplete data, and experimental study of link recommendation.
Conceptual modeling specifies the kinds of objects to be represented in an information system (IS). Conceptual models typically represent classes (categories, kinds) of objects rather than concrete specific individuals. While representation of classes may differ between grammars, a common design principle (DP) is what we term different semantics same syntax (D3S). Under this DP all classes are depicted using the same syntactic symbols. Following recent findings in psychology, we introduce a novel DP semantics-contingent syntax (SCS) whereby syntactic representations of classes in conceptual models may differ based on their meaning. We believe, SCS carries profound implications for theory and practice of conceptual modeling that we hope to explore in future work.
To appropriately use the information from social media to solve finance-related problems is challenging to both finance and data mining. Traditional schemes in finance focus on identifying the trading activities and financial events that cause asset price shocks, while the usage of data typically only covers standard events such as earning announcements, financial statements, and new stock issuance. Related data-driven implementations, on the other hand, mainly focus on developing trading strategies using social media data, while the results usually lack theoretical explanations. To fill the gap between the utilization of social media data and financial theories, we develop a Degree of Social Attention (DSA) framework, by leveraging on the vast social networks data, to bring profound impacts on research and practice in finance including market efficiency analysis. For each stock, the framework dynamically generates a DSA to capture its price shock by modeling the topological structure of the social network as well as the self-influence of each social media user. Based on a newly proposed influence propagation model, we estimate the market influence of the current DSA and the effects of historical ones on different stocks. Finally, we conduct comprehensive evaluations with the real-world data from social media and the stock market. We verify the essential relationship between social media activities and the stock market movement by testing the DSA Hypothesis we formalize. We also evaluate the practical value of our framework in solving a stock shocks detection problem. The results suggest that considering historical DSAs improve the models fitting performance for price shocks in terms of the statistical significance as well as the ranking accuracy.