A Comprehensive Social Matrix Factorization with Social Regularization for Recommendation Based on Implicit Similarity by Fusing Trust Relationships and Social Tags

Social relationships play an important role in improving the quality of recommender systems (RSs). A large number of experimental results show that social relationship-based recommendation methods alleviate the problems of data sparseness and cold start in RSs to some extent. However, since the social relationships between users are extremely sparse and complex, and it is difficult to obtain accurately user preference model, thus the performance of the recommendation system is affected by the existing social recommendation methods. In order to accurately model social relationships and improve recommendation quality, we use explicit social relationships such as user-item ratings, trust relationships and implicit social relationships such as social tags to mine potential interest preferences of users and propose an improved social recommendation method integrating trust relationship and social tags. The method map user features and item features to the shared feature space by using the above social relationship, respectively, and obtains user similarity and item similarity through potential feature vectors of users and items, and continuously trains them to obtain accurate similarity relationship to improve the recommendation performance. Experimental results demonstrate that our proposed approach achieves superior performance to the other social recommendation approaches.

User Generated Social Tags Versus Librarian Generated Subject Headings, A Comparative Study in the Domain of History

Social tagging allows users to assign any free-form keywords as tags to any digital resources through a decentralised way. Many information scientists find that there are similarities through their studies between usergenerated social tags and the librarian-generated subject headings for the libraries. The present study was conducted to identify the similarity and dissimilarity between user-generated social tags and librarian-generated subject terms of 1000 books in the domain of History. The study also conducted to identify whether social tags can replace controlled vocabularies. The study finds that only a small portion of terms overlaps with each other (3.54 % of social tags & 56.07 % of SLSH terms) and Spearman’s rank correlation proves that there is a good association between overlapping terms. Jaccard similarity coefficient highlights that users and the librarian use different terminologies (as J = 0.13, 0.12 & 0.11). Individual title wise comparison also defines that 90 per cent (88.4 %) of all book titles where users and the librarian use at least one common term. Users use the least subject & non-subject terms but use some personal tags for personal benefit whereas the librarian use only subject & non-subject terms. Matching with each book title clarifies that for describing resources users mostly use title based keywords (696) whereas the librarian use very little title based keywords (113). The study clearly defines that social tags can enhance the experience of library users. If it can be exploited properly it can complement to controlled vocabularies but can not replace the controlled vocabularies used for libraries a long time. Overall the study explicitly identifies the viability regarding the adoption of social tags into the library databases where the resources in the field of history will be accessed.

Personalized Mobile Video Recommendation Based on User Preference Modeling by Deep Features and Social Tags

With the explosive growth of mobile videos, helping users quickly and effectively find mobile videos of interest and further provide personalized recommendation services are the developing trends of mobile video applications. Mobile videos are characterized by their wide variety, single content, and short duration, and thus traditional personalized video recommendation methods cannot produce effective recommendation performance. Therefore, a personalized mobile video recommendation method is proposed based on user preference modeling by deep features and social tags. The main contribution of our work is three-fold: (1) deep features of mobile videos are extracted by an improved exponential linear units-3D convolutional neural network (ELU-3DCNN) for representing video content; (2) user preference is modeled by combining user preference for deep features with user preference for social tags that are respectively modeled by maximum likelihood estimation and exponential moving average method; (3) a personalized mobile video recommendation system based on user preference modeling is built after detecting key frames with a differential evolution optimization algorithm. Experiments on YouTube-8M dataset have shown that our method outperforms state-of-the-art methods in terms of both precision and recall of personalized mobile video recommendation.

Social Tags Versus LCSH Descriptors a Comparative Metadata Analysis in the Field of Economics

The concept of ‘social tagging’ has gained popularity nowadays due to the emergence of web 2.0 technologies. Those technologies led to the practice of associating metadata with digital resources among users through collaboratively or socially for self-information retrieval. Many researchers have opined that social tags can enhance the use of library collections. The present study was predominantly carried out to compare social tags collected from the LibraryThing website with Library of Congress Subject Heading (LCSH) descriptors collected from the Library of Congress Online Catalogue applied for thousand book titles in the field of Economics. The study also aimed to know whether social tags can be applied in the library database or not. The findings elucidate that users mostly use descriptors (47.39 %) as tags than expert’s usage of tags (12.77 %) as descriptors. Spearman’s correlation suggests that 75 per cent chance where tags and descriptors can be used simultaneously in overlapping terms. The Jaccard similarity coefficient identifies that users and experts use different terminologies to annotate the books. Users and experts use at least one common keyword for major book titles (908). Users mostly sought title based keywords but experts use mostly subject-based terminologies. The study further clarifies that social tags may be incorporated into the library databases but cannot replace LCSHs. The accessibility and usage of documents especially in the field of economics may be enhanced once the notion of social tags is incorporated with the library OPAC.

Capturing Associations and Sustainable Competitiveness of Brands from Social Tags

With the explosion of social media, consumers’ minds have become important assets in brand competitions. Determining a brand’s competitive structure based on consumers’ desires is particularly important to effectively establish a brand and maintain sustainable competitiveness. The traditional methods of determining brand competitiveness are costly and time-consuming. In this study, we propose an efficient, systematical, highly automated, and real-time method to determine brand competitiveness based on consumers’ brand associations with the brand’s social tags. Using a set of 45 brands in the automobile industry and around 50,000 social tags, we compared our brand competitiveness determination method with data provided by Interbrand and directly elicited survey data, finding a significant correlation and a better predictive power in consumers’ perceived brand competitiveness than the traditional method. Our proposed method enables managers to create and maintain sustainable brand advantages in consumers’ minds.