A Unified Framework for Providing Recommendations in Social Tagging Systems Based on Ternary Semantic Analysis

Social Tagging is the process by which many users add metadata in the form of keywords, to annotate and categorize items (songs, pictures, Web links, products, etc.). Social tagging systems (STSs) can provide three different types of recommendations: They can recommend 1) tags to users, based on what tags other users have used for the same items, 2) items to users, based on tags they have in common with other similar users, and 3) users with common social interest, based on common tags on similar items. However, users may have different interests for an item, and items may have multiple facets. In contrast to the current recommendation algorithms, our approach develops a unified framework to model the three types of entities that exist in a social tagging system: users, items, and tags. These data are modeled by a 3-order tensor, on which multiway latent semantic analysis and dimensionality reduction is performed using both the Higher Order Singular Value Decomposition (HOSVD) method and the Kernel-SVD smoothing technique. We perform experimental comparison of the proposed method against state-of-the-art recommendation algorithms with two real data sets (Last.fm and BibSonomy). Our results show significant improvements in terms of effectiveness measured through recall/precision.     

Understanding why users tag: A survey of tagging motivation literature and results from an empirical study

While recent progress has been achieved in understanding the structure and dynamics of social tagging systems, we know little about the underlying user motivations for tagging, and how they influence resulting folksonomies and tags. This paper addresses three issues related to this question. (1) What distinctions of user motivations are identified by previous research, and in what ways are the motivations of users amenable to quantitative analysis? (2) To what extent does tagging motivation vary across different social tagging systems? (3) How does variability in user motivation influence resulting tags and folksonomies? In this paper, we present measures to detect whether a tagger is primarily motivated by categorizing or describing resources, and apply these measures to datasets from seven different tagging systems. Our results show that (a) users’ motivation for tagging varies not only across, but also within tagging systems, and that (b) tag agreement among users who are motivated by categorizing resources is significantly lower than among users who are motivated by describing resources. Our findings are relevant for (1) the development of tag-based user interfaces, (2) the analysis of tag semantics and (3) the design of search algorithms for social tagging systems.     

Understanding why users tag: A survey of tagging motivation literature and results from an empirical study

While recent progress has been achieved in understanding the structure and dynamics of social tagging systems, we know little about the underlying user motivations for tagging, and how they influence resulting folksonomies and tags. This paper addresses three issues related to this question. (1) What distinctions of user motivations are identified by previous research, and in what ways are the motivations of users amenable to quantitative analysis? (2) To what extent does tagging motivation vary across different social tagging systems? (3) How does variability in user motivation influence resulting tags and folksonomies? In this paper, we present measures to detect whether a tagger is primarily motivated by categorizing or describing resources, and apply these measures to datasets from seven different tagging systems. Our results show that (a) users’ motivation for tagging varies not only across, but also within tagging systems, and that (b) tag agreement among users who are motivated by categorizing resources is significantly lower than among users who are motivated by describing resources. Our findings are relevant for (1) the development of tag-based user interfaces, (2) the analysis of tag semantics and (3) the design of search algorithms for social tagging systems.     

Folksonomy-based personalized search and ranking in social media services

In recent years, social Web users have been overwhelmed by the huge numbers of social media available. Consequentially, users have trouble finding social media suited to their needs. To help such users retrieve useful social media content, we propose a new model of tag-based personalized searches to enhance not only retrieval accuracy but also retrieval coverage. By leveraging social tagging as a preference indicator, we build two models: (i) a latent tag preference model that reflects how a certain user has assigned tags similar to a given tag and (ii) a latent tag annotation model that captures how users have tagged a certain tag to resources similar to a given resource. We then seamlessly map the tags onto items, depending on an individual user's query, to find the most desirable content relevant to the user's needs. Experimental results demonstrate that the proposed method significantly outperforms the state-of-the art algorithms and show our method's feasibility for personalized searches in social media services.     

Semantic similarity measures for enhancing information retrieval in folksonomies

Collaborative tagging systems, also known as folksonomies, enable a user to annotate various web resources with a free set of tags for sharing and searching purposes. Tags in a folksonomy reflect users’ collaborative cognition about information. Tags play an important role in a folksonomy as a means of indexing information to facilitate search and navigation of resources. However, the semantics of the tags, and therefore the semantics of the resources, are neither known nor explicitly stated. It is therefore difficult for users to find related resources due to the absence of a consistent semantic meaning among tags. The shortage of relevant tags increases data sparseness and decreases the rate of information extraction with respect to user queries. Defining semantic relationships between tags, resources, and users is an important research issue for the retrieval of related information from folksonomies. In this research, a method for finding semantic relationships among tags is proposed. The present study considers not only the pairwise relationships between tags, resources, and users, but also the relationships among all three. Experimental results using real datasets from Flickr and Del.icio.us show that the method proposed here is more effective than previous methods such as LCH, JCN, and LIN in finding semantic relationships among tags in a folksonomy.     

A framework for tag-aware recommender systems

In social tagging system, a user annotates a tag to an item. The tagging information is utilized in recommendation process. In this paper, we propose a hybrid item recommendation method to mitigate limitations of existing approaches and propose a recommendation framework for social tagging systems. The proposed framework consists of tag and item recommendations. Tag recommendation helps users annotate tags and enriches the dataset of a social tagging system. Item recommendation utilizes tags to recommend relevant items to users. We investigate association rule, bigram, tag expansion, and implicit trust relationship for providing tag and item recommendations on the framework. The experimental results show that the proposed hybrid item recommendation method generates more appropriate items than existing research studies on a real-world social tagging dataset.     

Detecting Communities in <Emphasis Type="Italic">K</Emphasis>-Partite <Emphasis Type="Italic">K</Emphasis>-Uniform (Hyper)Networks

In social tagging systems such as Delicious and Flickr,users collaboratively manage tags to annotate resources.Naturally,a social tagging system can be modeled as a (user,tag,resource) hypernetwork,where there are three different types of nodes,namely users,resources and tags,and each hyperedge has three end nodes,connecting a user,a resource and a tag that the user employs to annotate the resource.Then how can we automatically cluster related users,resources and tags,respectively? This is a problem of community detection in a 3-partite,3-uniform hypernetwork.More generally,given a K-partite K-uniform (hyper)network,where each (hyper)edge is a K-tuple composed of nodes of K different types,how can we automatically detect communities for nodes of different types? In this paper,by turning this problem into a problem of finding an efficient compression of the (hyper)network’s structure,we propose a quality function for measuring the goodness of partitions of a K-partite K-uniform (hyper)network into communities,and develop a fast community detection method based on optimization.Our method overcomes the limitations of state of the art techniques and has several desired properties such as comprehensive,parameter-free,and scalable.We compare our method with existing methods in both synthetic and real-world datasets.     

Collaborative filtering based on collaborative tagging for enhancing the quality of recommendation

We propose a collaborative filtering method to provide an enhanced recommendation quality derived from user-created tags. Collaborative tagging is employed as an approach in order to grasp and filter users’ preferences for items. In addition, we explore several advantages of collaborative tagging for data sparseness and a cold-start user. These applications are notable challenges in collaborative filtering. We present empirical experiments using a real dataset from del.icio.us. Experimental results show that the proposed algorithm offers significant advantages both in terms of improving the recommendation quality for sparse data and in dealing with cold-start users as compared to existing work.     

一种面向协作标签系统的图片检索聚类方法

为了更有效地进行图片检索,提出了一种面向Web2.0协作标签系统的图片检索聚类方法。该算法首先针对标签空间由于标签表达多样性带来的不一致问题,并通过挖掘标签间的词汇关系实现语义级查询扩展来得到语义可能相关的扩展图片结果集;然后根据标签间的相关度度量选出图片结果集中与查询标签高相关的标签集,接着采用一种自顶向下启发式的图划分算法来自动对次相关标签集进行分类。最后图片结果集即根据标签分类结果被聚类。为验证该方法的效果,从标签图片共享网站Flickr上随机下载了大量真实图片集以及所含带的标签元数据,在已实现的图片检索原型系统PivotBrowser上进行了大量实验,结果证明,该聚类算法能有效解决标签空间存在的标签表达不一致问题和标签查询歧义性问题,能提供更满意的用户检索。     

Bridging the gap between tagging and querying vocabularies: Analyses and applications for enhancing multimedia IR

Collaborative tagging has become an increasingly popular means for sharing and organizing Web resources, leading to a huge amount of user-generated metadata. These annotations represent quite a few different aspects of the resources they are attached to, but it is not obvious which characteristics of the objects are predominantly described. The usefulness of these tags for finding/re-finding the annotated resources is also not completely clear. Several studies have started to investigate these issues, however only by focusing on a single type of tagging system or resource. We study this problem across multiple domains and resource types and identify the gaps between the tag space and the querying vocabulary. Based on the findings of this analysis, we then try to bridge the identified gaps, focusing in particular on multimedia resources. We focus on the two scenarios of music and picture resources and develop algorithms, which identify usage (theme) and opinion (mood) characteristics of the items. The mood and theme labels our algorithms infer are recommended to the users, in order to support them during the annotation process. The evaluation of the proposed methods against user judgements, as well as against expert ground truth reveal the high quality of our recommended annotations and provide insights into possible extensions for music and picture tagging systems to support retrieval.     

Tag recommendation by machine learning with textual and social features

Tags are very popular in social media (like Youtube, Flickr) and provide valuable and crucial information for social media. But at the same time, there exist a great number of noisy tags, which lead to many studies on tag suggestion and recommendation for items including websites, photos, books, movies, and so on. The textual features of tags, likes tag frequency, have mostly been used in extracting tags that are related to items. In this paper, we address the problem of tag recommendation for social media users. This issue is as important as the tag recommendation for items, because the tags representing users are strongly related to the users’ favorite topics. We propose several novel features of tags for machine learning that we call social features as well as textual features. The experimental results of Flickr show that our proposed scheme achieves viable performance on tag recommendation for users.     

Summarizing dynamic Social Tagging Systems

Social tagging is a popular method that allows users of social networks to share annotation in the form of keywords, called tags, assigned to resources. Social tagging addresses information overload by easing the task of locating interesting entities in a social network. Nevertheless, users can still be overwhelmed by too many tags posted at each moment. A process is needed that offers an accurate overview of the representative entities and their relationships with each other, while dealing with the dynamics of social tagging and of tags’ semantics. We propose a method for the automated summarization of an evolving multi-modal social network, focusing on the entities that stay representative over time for some subnetwork in the social tagging system. We report on experiments with real data from the Bibsonomy social tagging system, where we compare our dynamic approach with a static one.     

采用对象特征向量表示法的标签聚类算法

在社会化标记系统中,常采用聚类等数据挖掘技术来解决标签冗余和语意模糊的问题.现有标签聚类算法大多根据不同标签在对象中共同出现的次数来计算它们之间的相似度,但是这种方法聚类的精确度与召回率并不高.针对此问题,提出一种新的标签聚类算法,充分考虑标签的标记信息,采用基于对象的特征向量来精确地表征一个标签,根据余弦相似度公式得到较为准确的标签相似度,然后采用K-Means算法将用户标签进行聚类.实验结果表明该算法能够得到更加精确的聚类结果.     

Ranking and Suggesting Popular Items

We consider the problem of ranking the popularity of items and suggesting popular items based on user feedback. User feedback is obtained by iteratively presenting a set of suggested items, and users selecting items based on their own preferences either from this suggestion set or from the set of all possible items. The goal is to quickly learn the true popularity ranking of items (unbiased by the made suggestions), and suggest true popular items. The difficulty is that making suggestions to users can reinforce popularity of some items and distort the resulting item ranking. The described problem of ranking and suggesting items arises in diverse applications including search query suggestions and tag suggestions for social tagging systems. We propose and study several algorithms for ranking and suggesting popular items, provide analytical results on their performance, and present numerical results obtained using the inferred popularity of tags from a month-long crawl of a popular social bookmarking service. Our results suggest that lightweight, randomized update rules that require no special configuration parameters provide good performance.     

Social media filtering based on collaborative tagging in semantic space

We propose a semantic collaborative filtering method to enhance recommendation quality derived from user-generated tags. Social tagging is employed as an approach in order to grasp and filter users’ preferences for items. In addition, we explore several advantages of semantic tagging for ambiguity, synonymy, and semantic interoperability, which are notable challenges in information filtering. The proposed approach first determines semantically similar users using social tagging and subsequently discovers semantically relevant items for each user. Experimental results show that our method offers significant advantages both in terms of improving the recommendation quality and in dealing with ambiguity, synonymy, and interoperability issues.     

Multi-label multi-instance learning with missing object tags

In this paper, a novel framework is developed for leveraging large-scale loosely tagged images for object classifier training by addressing three key issues jointly: (a) spam tags e.g., some tags are more related to popular query terms rather than the image semantics; (b) loose object tags, e.g., multiple object tags are loosely given at the image level without identifying the object locations in the images; (c) missing object tags, e.g., some object tags are missed and thus negative bags may contain positive instances. To address these three issues jointly, our framework consists of the following key components for leveraging large-scale loosely tagged images for object classifier training: (1) distributed image clustering and inter-cluster visual correlation analysis for handling the issue of spam tags by filtering out large amounts of junk images automatically, (2) multiple instance learning with missing tag prediction for dealing with the issues of loose object tags and missing object tags jointly; (3) structural learning for leveraging the inter-object visual correlations to train large numbers of inter-related object classifiers jointly. Our experiments on large-scale loosely tagged images have provided very positive results.     

Predicting the environment from social media: A collective classification approach

We propose a method which uses Flickr tags to predict a wide variety of environmental features, such as climate data, land cover categories, species occurrence, and human assessments of scenicness. The role of Flickr tags in our method is two-fold. First, we show that Flickr tags capture information which is highly complementary to what is found in traditional structured environmental datasets. By combining Flickr tags with traditional datasets, we can thus make more accurate predictions than is possible using either Flickr tags or traditional datasets alone. Second, we propose a collective prediction model which crucially relies on Flickr tags to define a neighbourhood structure. The use of a collective prediction formulation is motivated by the fact that most environmental features are strongly spatially autocorrelated. While this suggests that geographic distance should play a key role in determining neighbourhoods, we show that considerable gains can be made by additionally taking Flickr tags and traditional data into consideration.     

丰富图像标签的正则化非负矩阵分解方法

目的 随着Web2.0技术的进步,以用户生成内容为中心的社交网站蓬勃发展,也使得基于图像标签的图像检索技术越来越重要。但是,由于用户标注时的随意性和个性化,导致用户提交的图像标签不够完备,降低了图像检索的准确性。方法 针对这一问题,提出一种正则化的非负矩阵分解方法来丰富图像欠完备的标签,提高图像标签的完备性。利用非负矩阵分解的方法将原始的标签-图像矩阵投影到潜在的低秩空间里消除噪声,同时利用图像的类内视觉离散度作为正则化项提高消除噪声、丰富标签的效果。结果 利用从社交网站Flickr上下载的大量社交图像进行对比实验,验证了本文方法对丰富图像标签的有效性。通过对比目前流行的优化算法,本文算法获得较高的性能提升,算法平均准确度提高了12.3%。结论 将图像类内视觉离散度作为正则化项的非负矩阵分解算法,能较好地丰富社交图像的标签,解决网络图像标签的欠完备问题。     

Qualitative analysis of user-based and item-based prediction algorithms for recommendation agents

Recommendation agents employ prediction algorithms to provide users with items that match their interests. In this paper, several prediction algorithms are described and evaluated, some of which are novel in that they combine user-based and item-based similarity measures derived from either explicit or implicit ratings. Both statistical and decision-support accuracy metrics of the algorithms are compared against different levels of data sparsity and different operational thresholds. The first metric evaluates the accuracy in terms of average absolute deviation, while the second evaluates how effectively predictions help users to select high-quality items. The experimental results indicate better performance of item-based predictions derived from explicit ratings in relation to both metrics. Category-boosted predictions lead to slightly better predictions when combined with explicit ratings, while implicit ratings, in the context that have been defined in this paper, perform much worse than explicit ratings.