Effective and Efficient Query-aware Snippet Extraction for Web Search

Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 3035 - 3046 December 7-11, 2022 ©2022 Association for Computational Linguistics Effective and Efficient Query-aware Snippet Extraction for Web Search Jingwei Yi1, Fangzhao Wu2, Chuhan Wu3, Xiaolong Huang4, Binxing Jiao4, Guangzhong Sun1, Xing Xie2 1University of Science and Technology of China 2Microsoft Research Asia 3Tsinghua University 4Microsoft STC Asia yjw1029@mail.ustc.edu.cn {wufangzhao,wuchuhan15}@gmail.com {xiaolhu,binxjia,xingx}@microsoft.com gzsun@ustc.edu.cn Abstract Query-aware webpage snippet extraction is widely used in search engines to help users better understand the content of the returned webpages before clicking. Although important, it is very rarely studied. In this paper, we pro- pose an effective query-aware webpage snippet extraction method named DeepQSE, aiming to select a few sentences which can best summa- rize the webpage content in the context of input query. DeepQSE first learns query-aware sen- tence representations for each sentence to cap- ture the fine-grained relevance between query and sentence, and then learns document-aware query-sentence relevance representations for snippet extraction. Since the query and each sentence are jointly modeled in DeepQSE, its online inference may be slow. Thus, we fur- ther propose an efficient version of DeepQSE, named Efficient-DeepQSE, which can signif- icantly improve the inference speed of Deep- QSE without affecting its performance. The core idea of Efficient-DeepQSE is to decom- pose the query-aware snippet extraction task into two stages, i.e., a coarse-grained candidate sentence selection stage where sentence repre- sentations can be cached, and a fine-grained relevance modeling stage. Experiments on two real-world datasets validate the effectiveness and efficiency of our methods. 1 Introduction Given an input search query, search engines such as Google 1 and Bing 2, not only return the URLs and the titles of the relevant webpages, but also show the query-aware snippets of these webpages, aiming to help users better understand the webpage content before clicking. These webpage snippets are usually one or two sentences extracted from the webpage, which can not only summarize the key content of the webpage, but also be relevant 1https://www.google.com 2https://www.bing.com einstein achievement Query einstein Query Title Title Snippet Snippet https://en.Wikipedia.org > wiki > Albert_Einstein Albert Einstein - Wikipedia Listen); 14 March 1879 – 18 April 1955) was a German-born theoretical physicist, widely Acknowledged to be one of the greatest physicists of all time, Einstein … Nationality: Germany, US, Swiss, Austro-Hung… Date of death: 18 April 1955 https://en.Wikipedia.org > wiki > Albert_Einstein Albert Einstein - Wikipedia His work is also known for its influence on the philosophy of science. He received the 1921 Nobel Price in Physics “for his services to the theoretical physics, and … Figure 1: Examples of query-aware snippets in search engines. to the input query. Some examples are shown in Figure 1. For the query ‘einstein’ and the webpage of ‘Albert Einstein - Wikipedia’, a good snippet is a brief introduction of Einstein’s life. While for the query ‘einstein achievement’, a good snippet would be sentences describing his influence on science. In other words, the snippet is a summarization of the webpage in …

Content truncated to 500 words