Upper Bound Approximation for BlockMaxWand

Macdonald, Craig; Tonellotto, Nicola

doi:10.1145/3121050.3121094

Cited by 4 publications

(2 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With regular document-ordered indexes, only term-frequencies are stored within the postings lists, which means that the ranker (or parameters of the ranker) can be easily changed without recomputing the entire index. Even with dynamic pruning methods, such as WAND and BMW, score upper-bounds can be estimated independently from the ranker, allowing the ranker to be changed efficiently [159,160]. Future work should investigate these operational trade-offs that are important for real IR systems in practice.…”

Section: Discussionmentioning

confidence: 99%

Managing tail latency in large scale information retrieval systems

Mackenzie

2020

SIGIR Forum

View full text Add to dashboard Cite

As both the availability of internet access and the prominence of smart devices continue to increase, data is being generated at a rate faster than ever before. This massive increase in data production comes with many challenges, including efficiency concerns for the storage and retrieval of such large-scale data. However, users have grown to expect the sub-second response times that are common in most modern search engines, creating a problem --- how can such large amounts of data continue to be served efficiently enough to satisfy end users? This dissertation investigates several issues regarding tail latency in large-scale information retrieval systems. Tail latency corresponds to the high percentile latency that is observed from a system --- in the case of search, this latency typically corresponds to how long it takes for a query to be processed. In particular, keeping tail latency as low as possible translates to a good experience for all users, as tail latency is directly related to the worst-case latency and hence, the worst possible user experience. The key idea in targeting tail latency is to move from questions such as "what is the median latency of our search engine?" to questions which more accurately capture user experience such as "how many queries take more than 200 ms to return answers?" or "what is the worst case latency that a user may be subject to, and how often might it occur?" While various strategies exist for efficiently processing queries over large textual corpora, prior research has focused almost entirely on improvements to the average processing time or cost of search systems. As a first contribution, we examine some state-of-the-art retrieval algorithms for two popular index organizations, and discuss the trade-offs between them, paying special attention to the notion of tail latency. This research uncovers a number of observations that are subsequently leveraged for improved search efficiency and effectiveness. We then propose and solve a new problem, which involves processing a number of related query variations together, known as multi-queries , to yield higher quality search results. We experiment with a number of algorithmic approaches to efficiently process these multi-queries, and report on the cost, efficiency, and effectiveness trade-offs present with each. Finally, we examine how predictive models can be used to improve the tail latency and end-to-end cost of a commonly used multi-stage retrieval architecture without impacting result effectiveness. By combining ideas from numerous areas of information retrieval, we propose a prediction framework which can be used for training and evaluating several efficiency/effectiveness trade-off parameters, resulting in improved trade-offs between cost, result quality, and tail latency.

show abstract

Section: Discussionmentioning

confidence: 99%

Managing tail latency in large scale information retrieval systems

Mackenzie

2020

SIGIR Forum

View full text Add to dashboard Cite

show abstract

“…An interesting coda to this story is that academic researchers were exploring alternatives to per-block impact scores circa 2017, for exactly the same reason (to allow the scoring model to be defined at search time). For example, Macdonald and Tonellotto [10] showed how to derive tight approximate upper bounds for block-max Wand, based on work that dates back to 2011 [9]. Similarly, the recently-released PISA research system stores flexible block-level metadata [11].…”

Section: Setting the Stagementioning

confidence: 99%

From MAXSCORE to Block-Max Wand: The Story of How Lucene Significantly Improved Query Evaluation Performance

Grand¹,

Muir²,

Ferenczi³

et al. 2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The latest major release of Lucene (version 8) in March 2019 incorporates block-max indexes and exploits the block-max variant of Wand for query evaluation, which are innovations that originated from academia. This paper shares the story of how this came to be, which provides an interesting case study at the intersection of reproducibility and academic research achieving impact in the "real world". We offer additional thoughts on the often idiosyncratic processes by which academic research makes its way into deployed solutions.

show abstract

Examining the Additivity of Top-k Query Processing Innovations

Mackenzie

Moffat

2020

Proceedings of the 29th ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

Upper Bound Approximation for BlockMaxWand

Cited by 4 publications

References 11 publications

Managing tail latency in large scale information retrieval systems

Managing tail latency in large scale information retrieval systems

From MAXSCORE to Block-Max Wand: The Story of How Lucene Significantly Improved Query Evaluation Performance

Examining the Additivity of Top-k Query Processing Innovations

Contact Info

Product

Resources

About