Efficient Tag Recommendation for Real-Life Data

Abstract

Despite all of the advantages of tags as an easy and flexible information management approach, tagging is a cumbersome task. A set of descriptive tags has to be manually entered by users whenever they post a resource. This process can be simplified by the use of tag recommendation systems. Their objective is to suggest potentially useful tags to the user. We present a hybrid tag recommendation system together with a scalable, highly efficient system architecture. The system is able to utilize user feedback to tune its parameters to specific characteristics of the underlying tagging system and adapt the recommendation models to newly added content. The evaluation of the system on six real-life datasets demonstrated the system’s ability to combine tags from various sources (e.g., resource content or tags previously used by the user) to achieve the best quality of recommended tags. It also confirmed the importance of parameter tuning and content adaptation. A series of additional experiments allowed us to better understand the characteristics of the system and tagging datasets and to determine the potential areas for further system development.

The impact of image descriptions on user tagging behavior: A study of the nature and functionality of crowdsourced tags

“…Most notable here are the studies of Heyman et al (2008) and Lipczak and Milios (2011) Imitation models focused on single user behavior predict that words from descriptions will complement the terms produced by observing images making the resulting tags more diverse.…”

Extending FolkRank with content data

“…In content-based approaches, the textual content of the documents is used for either tag extraction and expansion [7] [8], or with document classification techniques [3] [11]. The effectiveness of tag recommendation algorithms is usually evaluated against simple baseline recommenders.…”

“…Few approaches address this new item problem, concentrating instead on artificially created post-core datasets where it is guaranteed that all users and documents in the test set are known to the system and already have some tags assigned to them. An exception is the hybrid recommender presented by Lipczak et al [7] [8] which utilises content data and can recommend tags for new documents and users. In order to recommend tags for new documents, approaches are required which model documents not only based on the tags assigned to them in the past (if any), but also their content.…”

Topic diversity in tag recommendation

“…The tag recommendation literature is rich in techniques exploiting co-occurrence patterns computed over a history of tag assignments [5,4,27], words extracted from multiple textual features of the target object [18], as well as metrics of tag relevance to filter out irrelevant terms or give more importance to the relevant ones [5,18]. In [5], we proposed several heuristic methods that jointly exploit these three dimensions, and showed that they outperform previous approaches in various datasets.…”

“…Tag recommendation methods have historically focused mostly on maximizing the relevance of the recommended tags [5,18,27]. When recommending tags for a target object, relevance refers to how well the recommended tags describe the contents of the target object.…”