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 2212.05189cs.LGcs.AI
收藏

扩展知识图谱,让人参与其中

Expanding Knowledge Graphs with Humans in the Loop

Cornell University(康奈尔大学);University of Wisconsin Madison(威斯康星大学麦迪逊分校);Pinterest

Emaad Manzoor, Jordan Tong, Sriniketh Vijayaraghavan, Rui Li

摘要:

策划的知识图谱编码领域专业知识,并提高多个领域中的推荐、细分、广告定位和其他机器学习系统的性能。当新概念出现在某个领域时,必须扩展知识图谱以保持机器学习性能。然而,手动扩展知识图谱在规模上是不可行的。在这项工作中,我们提出了一种方法,知识图谱扩展与人类在循环。具体地说,给定一个知识图谱,我们的方法预测新概念的“父母”被添加到这个图中,供人类专家进一步验证。我们表明,我们的方法是准确的和可证明的“人性化”。具体来说,我们证明了我们的方法预测的父母是“附近”的概念的真正的父母在知识图谱,即使预测是不正确的。然后,我们通过一个对照实验表明,满足这一特性可以提高人类与算法协作的速度和准确性。我们进一步评估了我们的方法在Pinterest的知识图谱上,并表明它在准确性和人性化方面优于竞争方法。在Pinterest的生产中部署后,我们的方法将知识图谱扩展所需的时间减少了约400%(与手动扩展相比),并为随后的广告收入增加了20%。

英文摘要:

Curated knowledge graphs encode domain expertise and improve the performance of recommendation, segmentation, ad targeting, and other machine learning systems in several domains. As new concepts emerge in a domain, knowledge graphs must be expanded to preserve machine learning performance. Manually expanding knowledge graphs, however, is infeasible at scale. In this work, we propose a method for knowledge graph expansion with humans-in-the-loop. Concretely, given a knowledge graph, our method predicts the "parents" of new concepts to be added to this graph for further verification by human experts. We show that our method is both accurate and provably "human-friendly". Specifically, we prove that our method predicts parents that are "near" concepts' true parents in the knowledge graph, even when the predictions are incorrect. We then show, with a controlled experiment, that satisfying this property increases both the speed and the accuracy of the human-algorithm collaboration. We further evaluate our method on a knowledge graph from Pinterest and show that it outperforms competing methods on both accuracy and human-friendliness. Upon deployment in production at Pinterest, our method reduced the time needed for knowledge graph expansion by ~400% (compared to manual expansion), and contributed to a subsequent increase in ad revenue of 20%.

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。

↑