Matches in ScholarlyData for { ?s ?p ?o. }
- item-1 next item-2.
- item-2 type ListItem.
- item-2 hasContent vahab-mirrokni.
- item-2 next item-3.
- item-2 previous item-1.
- item-3 type ListItem.
- item-3 hasContent s-muthukrishnan.
- item-3 previous item-2.
- 829 creator b-barla-cambazoglu.
- 829 creator baoqiu-cui.
- 829 creator flavio-junqueira.
- 829 creator scott-banachowski.
- 829 creator swee-lim.
- 829 creator vassilis-plachouras.
- 829 type InProceedings.
- 829 label "A refreshing perspective of search engine caching".
- 829 sameAs 829.
- 829 abstract "Web search engines are required to process very large amounts of data with a very low latency to return the best answers to user queries. To achieve low latency, designers of search engines take advantage of the query frequency distribution and employ caching at various levels of the system, e.g., caching of query results. The search engine also needs to update its index frequently to incorporate changes to the Web. After every index update, however, the query results for some entries in the cache might be stale. One solution to this problem is to flush the content of the cache. As the frequency of index updates increases, this solution imposes a significant performance penalty. In this work, we first argue that eviction policies are not as important as policies for invalidating queries and updating the cache content upon new versions of the index. We then introduce a novel caching algorithm that uses a time-to-live value to expire queries and refreshes queries using idle cycles of the engine. The cache selects the queries to refresh according to their frequency or temperature and to their age. More frequent and older queries are selected for refreshing because they are more likely to result in more hits in the future. In addition, we present mechanisms for setting the rate of refreshes issued by the cache. Our evaluation results on real workloads show that our cache outperforms an extension of LRU with refreshes, while the refresh rate is set so that the latency of queries is within a specified range. An implementation of this algorithm is currently in production use.".
- 829 hasAuthorList authorList.
- 829 isPartOf proceedings.
- 829 keyword "Indexing".
- 829 keyword "caching".
- 829 keyword "distribution".
- 829 keyword "index compression".
- 829 title "A refreshing perspective of search engine caching".
- authorList type List.
- authorList hasFirstItem item-1.
- authorList hasItem item-1.
- authorList hasItem item-2.
- authorList hasItem item-3.
- authorList hasItem item-4.
- authorList hasItem item-5.
- authorList hasItem item-6.
- authorList hasLastItem item-6.
- item-1 type ListItem.
- item-1 hasContent flavio-junqueira.
- item-1 next item-2.
- item-2 type ListItem.
- item-2 hasContent b-barla-cambazoglu.
- item-2 next item-3.
- item-2 previous item-1.
- item-3 type ListItem.
- item-3 hasContent vassilis-plachouras.
- item-3 next item-4.
- item-3 previous item-2.
- item-4 type ListItem.
- item-4 hasContent swee-lim.
- item-4 next item-5.
- item-4 previous item-3.
- item-5 type ListItem.
- item-5 hasContent baoqiu-cui.
- item-5 next item-6.
- item-5 previous item-4.
- item-6 type ListItem.
- item-6 hasContent scott-banachowski.
- item-6 previous item-5.
- 83 creator chihung-chi.
- 83 creator guillaume-pierre.
- 83 creator jiang-dejun.
- 83 type InProceedings.
- 83 label "Autonomous Resource Provisioning for Multi-Service Web Applications".
- 83 sameAs 83.
- 83 abstract "Dynamic resource provisioning aims at maintaining the end-to-end response time of a web application within a pre-defined SLA. Although the topic has been well studied for monolithic applications, provisioning resources for applications composed of multiple services remains a challenge. When the SLA is violated, one must decide which service(s) should be reprovisioned for optimal effect. We propose to assign an SLA only to the front-end service. Other services are not given any particular response time objectives. Services are autonomously responsible for their own provisioning operations and collaboratively negotiate performance objectives with each other to decide the provisioning service(s). We demonstrate through extensive experiments that our system can add/remove/shift both servers and caches within an entire multi-service application under varying workloads to meet the SLA target and improve resource utilization.".
- 83 hasAuthorList authorList.
- 83 isPartOf proceedings.
- 83 keyword "Capacity planning".
- 83 keyword "resource provisioning".
- 83 title "Autonomous Resource Provisioning for Multi-Service Web Applications".
- authorList type List.
- authorList hasFirstItem item-1.
- authorList hasItem item-1.
- authorList hasItem item-2.
- authorList hasItem item-3.
- authorList hasLastItem item-3.
- item-1 type ListItem.
- item-1 hasContent jiang-dejun.
- item-1 next item-2.
- item-2 type ListItem.
- item-2 hasContent guillaume-pierre.
- item-2 next item-3.
- item-2 previous item-1.
- item-3 type ListItem.
- item-3 hasContent chihung-chi.
- item-3 previous item-2.
- 842 creator fan-li.
- 842 creator hongyuan-zha.
- 842 creator jiang-bian.
- 842 creator xin-li.
- 842 creator zhaohui-zheng.
- 842 type InProceedings.
- 842 label "Ranking Specialization for Web Search: A Divide-and-Conquer Approach by Using Topical RankSVM".
- 842 sameAs 842.
- 842 abstract "Many ranking algorithms applying machine learning techniques have been proposed in informational retrieval and Web search. However, most of existing approaches do not explicitly take into account the fact that queries vary significantly in terms of ranking and entail different treatments regarding the ranking models. In this paper, we apply a divide-and-conquer framework for ranking specialization, i.e. learning multiple ranking models by addressing query difference. We first generate query representation by aggregating ranking features through pseudo feedback, and employ unsupervised clustering methods to identify a set of ranking-sensitive query topics based on training queries. To learn multiple ranking models for respective ranking-sensitive query topics, we define a global loss function by combining the ranking risks of all query topics, and we propose a unified SVM-based learning process to minimize the global loss. Moreover, we employ an ensemble approach to generate the ranking result for each test query by applying a set of ranking models of the most appropriate query topics. We conduct experiments using a benchmark dataset for learning ranking functions as well as a dataset from a commercial search engine. Experimental results show that our proposed approach can significantly improve the ranking performance over existing single-model approaches as well as straightforward local ranking approaches, and the automatically identified ranking-sensitive topics are more useful for enhancing ranking performance than pre-defined query categorization.".
- 842 hasAuthorList authorList.
- 842 isPartOf proceedings.
- 842 keyword "Query processing".
- 842 keyword "diversity".
- 842 keyword "ranking".
- 842 keyword "relevance".
- 842 keyword "retrieval".