arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1809.11084cs.DBcs.LGstat.ML
收藏

基于迁移学习的实体解析重用与自适应

Reuse and Adaptation for Entity Resolution through Transfer Learning

Qatar Computing Research Institute, HBKU(HBKU卡塔尔计算研究所);Nanyang Technological University(南洋理工大学)

Saravanan Thirumuruganathan, Shameem A Puthiya Parambath, Mourad Ouzzani, Nan Tang, Shafiq Joty

摘要:

实体解析是数据集成中的基本问题之一,基于机器学习的分类器通常可以提供最先进的结果。相当多的人力投入到特征工程和训练数据创建中。本文研究了一个新的问题:给定一个用于ER的数据集D_T,其训练数据有限或没有训练数据,是否有可能通过重用和适应来自相同或相关领域的数据集D_S的训练数据,在D_T上训练出一个好的ML分类器?我们的主要贡献包括:(1)一个分布式表示为基础的方法来编码每个元组从不同的数据集到一个标准的特征空间;(2)识别常见的情况下,重用的训练数据可以是有益的;和(3)五个算法处理上述情况。我们对来自5个不同领域(出版物,电影,歌曲,餐馆和书籍)的12个数据集进行了全面的实验。我们的实验表明,我们的算法提供了显着的好处,如提供优良的性能为固定的训练数据大小。

英文摘要:

Entity resolution (ER) is one of the fundamental problems in data integration, where machine learning (ML) based classifiers often provide the state-of-the-art results. Considerable human effort goes into feature engineering and training data creation. In this paper, we investigate a new problem: Given a dataset D_T for ER with limited or no training data, is it possible to train a good ML classifier on D_T by reusing and adapting the training data of dataset D_S from same or related domain? Our major contributions include (1) a distributed representation based approach to encode each tuple from diverse datasets into a standard feature space; (2) identification of common scenarios where the reuse of training data can be beneficial; and (3) five algorithms for handling each of the aforementioned scenarios. We have performed comprehensive experiments on 12 datasets from 5 different domains (publications, movies, songs, restaurants, and books). Our experiments show that our algorithms provide significant benefits such as providing superior performance for a fixed training data size.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑