Learning to rank for information retrieval and natural language processing [electronic resource]

• 作者： Li, Hang, 1965- author.
• 其他題名：
  • Synthesis lectures on human language technologies,
• 出版： San Rafael, California : Springer International Publishing AG
• 版本：Second edition.
• 叢書名： Synthesis lectures on human language technologies,#26
• 主題： Ranking and selection (Statistics) , Information retrieval , Natural language processing (Computer science) , Machine learning , Electronic books.
• ISBN： 9783031021558 (ebook) 、 9781627055857 (ebook) 、 1627055851 (ebook) 、 9783031010279 、 1627055843 、 9781627055840
• FIND@SFXID: CGU
• 資料類型: 電子書
• 內容註: Includes bibliographical references (pages 93-105). 1. Learning to rank -- 1.1 Ranking -- 1.2 Learning to rank -- 1.3 Ranking creation -- 1.4 Ranking aggregation -- 1.5 Learning for ranking creation -- 1.6 Learning for ranking aggregation -- 2. Learning for ranking creation -- 2.1 Document retrieval as example -- 2.2 Learning task -- 2.2.1 Training and testing -- 2.2.2 Training data creation -- 2.2.3 Feature construction -- 2.2.4 Evaluation -- 2.2.5 Relations with other learning tasks -- 2.3 Learning approaches -- 2.3.1 Pointwise approach -- 2.3.2 Pairwise approach -- 2.3.3 Listwise approach -- 2.3.4 Evaluation results -- 3. Learning for ranking aggregation -- 3.1 Learning task -- 3.2 Learning methods -- 4. Methods of learning to rank -- 4.1 PRank -- 4.1.1 Model -- 4.1.2 Learning algorithm -- 4.2 OC SVM -- 4.2.1 Model -- 4.2.2 Learning algorithm -- 4.3 McRank -- 4.3.1 Model -- 4.3.2 Learning algorithm -- 4.4 Ranking SVM -- 4.4.1 Linear model as ranking function -- 4.4.2 Ranking SVM model -- 4.4.3 Learning algorithm -- 4.5 IR SVM -- 4.5.1 Modified loss function -- 4.5.2 Learning algorithm -- 4.6 GBRank -- 4.6.1 Loss function -- 4.6.2 Learning algorithm -- 4.7 RankNet -- 4.7.1 Loss function -- 4.7.2 Model -- 4.7.3 Learning algorithm -- 4.7.4 Speed up of training -- 4.8 ListNet and ListMLE -- 4.8.1 Plackett-Luce model -- 4.8.2 ListNet -- 4.8.3 ListMLE -- 4.9 AdaRank -- 4.9.1 Loss function -- 4.9.2 Learning algorithm -- 4.10 SVM MAP -- 4.10.1 Loss function -- 4.10.2 Learning algorithms -- 4.11 SoftRanK -- 4.11.1 Soft NDCG -- 4.11.2 Approximation of rank distribution -- 4.11.3 Learning algorithm -- 4.12 LambdaRank -- 4.12.1 Loss function -- 4.12.2 Learning algorithm -- 4.13 LambdaMART -- 4.13.1 Model and loss function -- 4.13.2 Learning algorithm -- 4.14 Borda count -- 4.15 Markov chain -- 4.16 Cranking -- 4.16.1 Model -- 4.16.2 Learning algorithm -- 4.16.3 Prediction -- 5. Applications of learning to rank -- 6. Theory of learning to rank -- 6.1 Statistical learning formulation -- 6.2 Loss functions -- 6.3 Relations between loss functions -- 6.4 Theoretical analysis -- 7. Ongoing and future work -- Bibliography -- Author's biography.
• More
• 摘要註: Learning to rank refers to machine learning techniques for training a model in a ranking task. Learning to rank is useful for many applications in information retrieval, natural language processing, and data mining. Intensive studies have been conducted on its problems recently, and significant progress has been made. This lecture gives an introduction to the area including the fundamental problems, major approaches, theories, applications, and future work. The author begins by showing that various ranking problems in information retrieval and natural language processing can be formalized as two basic ranking tasks, namely ranking creation (or simply ranking) and ranking aggregation. In ranking creation, given a request, one wants to generate a ranking list of offerings based on the features derived from the request and the offerings. In ranking aggregation, given a request, as well as a number of ranking lists of offerings, one wants to generate a new ranking list of the offerings.
• 讀者標籤：

  需登入
• 引用連結：

  Copy Link
• 目前引用數
• Share: