无论是手动还是电子形式的文件注释,其主要目的是允许那些可能无法控制原始文件的人提供对信息源的个人看法。除了为原始信息源提供个人评估之外,我们还在研究一种情况,即所做的注释可以用作文档跟踪和推荐服务的额外信息源。今天存在的大多数注释工具都是为独立使用而设计的,没有参考注释的创建者。我们提出了一个可以辅助文档跟踪和推荐服务的标注模型AMIEDoT(Annotation Model for Information Exchange and Document Tracking)。该模型基于注释行为中的三个参数。我们相信,将文档参数、时间和注释创建者的参数引入注释过程可以成为了解谁使用了文档、何时使用文档以及文档用于什么的可靠来源。除了文档跟踪,我们的模型不仅可以用于信息的选择性传播,还可以用于推荐服务。AMIEDoT还可用于信息共享和信息重用。
英文摘要:
The primary objective of document annotation in whatever form, manual or electronic is to allow those who may not have control to original document to provide personal view on information source. Beyond providing personal assessment to original information sources, we are looking at a situation where annotation made can be used as additional source of information for document tracking and recommendation service. Most of the annotation tools existing today were conceived for their independent use with no reference to the creator of the annotation. We propose AMIEDoT (Annotation Model for Information Exchange and Document Tracking) an annotation model that can assist in document tracking and recommendation service. The model is based on three parameters in the acts of annotation. We believe that introducing document parameters, time and the parameters of the creator of annotation into an annotation process can be a dependable source to know, who used a document, when a document was used and for what a document was used for. Beyond document tracking, our model can be used in not only for selective dissemination of information but for recommendation services. AMIEDoT can also be used for information sharing and information reuse.
Evaluating Document Representations for Content-based Legal Literature Recommendations
Open Legal Data(开放法律数据);German Research Center for Artificial Intelligence(德国人工智能研究中心);ETH Zurich(苏黎世联邦理工学院);University of Wuppertal(伍珀塔尔大学)
推荐系统帮助法律专业人士找到相关文献来支持他们的案件。尽管它的重要性的专业,法律应用并没有反映在推荐系统和表示学习研究的最新进展。同时,法律推荐系统通常在没有任何公共基准数据集的情况下在小规模用户研究中进行评估。因此,这些研究的重现性有限。为了解决研究与实践之间的差距,我们探索了一套国家的最先进的文件表示方法检索语义相关的美国判例法的任务。我们评估基于文本的(例如,fastText、Transformers)、基于引用的(例如,DeepWalk,Poincaré)和混合方法。我们使用两种银标准对2,964份文件进行了注释,共比较了27种方法。银标准是从Open Case Book和Wikispecies新创建的,可以在开放许可下重复使用,以促进可重复性。我们的实验表明,来自平均fastText单词向量(在法律语料库上训练)的文档表示产生了最好的结果,紧随其后的是庞加莱引用嵌入。以混合方式组合fastText和Poincaré进一步改善了整体结果。除了整体性能外,我们还根据文档长度,引用计数和建议的覆盖范围分析了这些方法。我们在https://github.com/malteos/legal-document-similarity/上公开我们的源代码、模型和数据集。
Enhancing Academic Paper Recommendations Using Fine-Grained Knowledge Entities and Multifaceted Document Embeddings
Department of Information Management(信息管理系);Nanjing University of Science & Technology(南京理工大学);School of Computer Engineering(计算机工程学院);Jiangsu University of Technology(江苏科技大学);School of Information Management(信息管理学院);Central China Normal University(中央财经大学)
Multi-criteria recommendation systems to foster online grocery
College of Computing and Information Technology, Arab Academy for Science, Technology and Maritime Transport (AASTMT)(阿拉伯科学、技术与海运学院(AASTMT)计算与信息技术学院);atlanTTic, Universidade de Vigo(维戈大学atlanTTic研究中心)
Multimodal and Multiscale Spatial-Temporal Semantic Search and Recommendation with AI Foundation Models
School of Geographical Sciences and Urban Planning, Arizona State University(地理科学与城市规划学院,亚利桑那州立大学);Alaska Native Tribal Health Consortium(阿拉斯加原住民部落健康联盟);Woodwell Climate Research Center(伍德沃德气候研究中心);School of Computing and Augmented Intelligence, Arizona State University(计算与增强智能学院,亚利桑那州立大学)
A Survey of Generative Search and Recommendation in the Era of Large Language Models
The Hong Kong Polytechnic University(香港理工大学);National University of Singapore(新加坡国立大学);University of Science and Technology of China(中国科学技术大学);Chinese Academy of Sciences(中国科学院);Harbin Institute of Technology (Shenzhen)(哈尔滨工业大学(深圳))