Estimating translation probabilities for social tag suggestion
作者：
Highlights：
• We present a new perspective to tag suggestion and treat it as a translation process.
• We propose two methods to estimate the translation probabilities.
• Our methods can solve the problem of vocabulary gap.
• Our methods are effective and robust compared with other methods.
• Our methods are relatively simple and efficient, which makes them practical.
摘要
•We present a new perspective to tag suggestion and treat it as a translation process.•We propose two methods to estimate the translation probabilities.•Our methods can solve the problem of vocabulary gap.•Our methods are effective and robust compared with other methods.•Our methods are relatively simple and efficient, which makes them practical.
论文关键词：Natural language processing,Tag suggestion,Translation model,Word alignment model,Pointwise mutual information
论文评审过程：Available online 12 October 2014.
论文官网地址：https://doi.org/10.1016/j.eswa.2014.10.002
