Diversifying with Few Regrets, But too Few to Mention

Hussain, Zaeem; Khan, H.; Sharaf, Mohamed A.

doi:10.1145/2795218.2795225

Cited by 8 publications

(9 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To address that limitation, in this work we posit that employing diversification techniques (e.g., [16], [26], [27], [28], [29], [30], [31], [32]) in the process of view recommendation allows eliminating that redundancy and provides a good and concise coverage of the possible insights to be discovered.…”

Section: ) Context-driven Similaritymentioning

confidence: 99%

Efficient Diversification for Recommending Aggregate Data Visualizations

2023

Self Cite

View full text Add to dashboard Cite

Visual data exploration is ubiquitous in nearly every industry and organization to support discovering data-driven actionable insights. However, unlocking those insights requires analysts to manually construct a prohibitively large number of aggregate queries and visually explore their results, looking for those valuable and insightful visualizations. Such a challenge naturally motivated the development of novel solutions that automate the visual exploration process, and recommend to analysts those particular queries that best visualize their data and reveal interesting actionable insights. In such automated solutions, there is a clear need for providing analysts with a diversified and concise set of recommended visualizations, which cover and represent a large combinatorial high-dimensional space of possible visualizations. However, directly incorporating existing diversification methods leads to a ''process-first-diversify-next'' approach, in which all possible data visualizations are generated first through executing a large number of aggregate queries. To address this challenge and minimize the incurred query processing costs, in this work, we propose novel optimization techniques for the efficient diversification of recommended insightful visualizations. The key idea underlying our proposed techniques is to identify and eliminate the processing of a large number of low-utility insignificant visualizations. Meanwhile, for the potentially high-utility insightful visualizations, shared multi-query optimization techniques are proposed for further reduction in data processing cost. Our extensive experimental evaluation on real datasets demonstrates the performance gains provided by our proposed techniques, in terms of minimizing the query processing cost (i.e., efficiency), as well as maximizing the quality of recommendations (i.e., effectiveness).

show abstract

Section: ) Context-driven Similaritymentioning

confidence: 99%

Efficient Diversification for Recommending Aggregate Data Visualizations

2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…The purpose of ReDi [30] is to combine the principle of regret minimization with the one of diversity maximization; these two approaches share the objective of identifying a representative set of tuples among the full dataset, even though they serve very different needs: the former aims at determining a subset of data that satisfies the user the most, the latter at providing data that has good coverage on the dataset and low redundancy. Nevertheless, as the authors point out, typically the user is able to provide query preferences only for a subset of attributes, expecting to get some more insights about the attributes that have not been taken into consideration from the result itself: hence the need of capturing diverse objects that summarize the unsolicited part of information that is stored in the dataset.…”

Section: Redimentioning

confidence: 99%

“…REGRET MINIMIZATION QUERIES k-Representative regret minimization [25] k-Regret Minimizing Sets [28] ReDi [30] Table 1: Comparison of strategies used to support preferential query answering rely on a quantitative representation of the user preferences, be it a class of scoring functions or a region surrounding a preference vector. Few other techniques instead make use of a qualitative representation using preference formulas (P-Skylines, Trade-off Skylines) or preference dimensions sets (SKYRANK).…”

Section: Preferential Query Answeringmentioning

confidence: 99%

“…Representative identification Restriction to attribute subspaces FLEXIBLE SKYLINES R-Skylines [9] Uncertain top-k queries [12] P-Skylines [13] Trade-off Skylines [16] SKYLINE RANKING Skyline ordering [18] k-Representative Skylines [19] SKYRANK [21] REGRET MINIMIZATION QUERIES k-Representative regret minimization [25] k-Regret Minimizing Sets [28] ReDi [30] Table 2: Comparison of strategies used to reduce query output size restricting the dominance check to a specific region around the user's preference vector, as shown in Figure 1: in this case the result size cannot be specified a priori, although it is worth noting that the more user's preference information is used in the query process, the more the output set is expected to decrease in size.…”

Section: User Preferences Integrationmentioning

confidence: 99%

See 1 more Smart Citation

Skyline Operators and Regret Minimization Techniques for Managing User Preferences in the Query Process

Talarico¹

2022

Preprint

View full text Add to dashboard Cite

The problem of selecting the most representative tuples from a dataset has led to the development of powerful tools, among which Skyline and Ranking (or Top-k) queries stand out for their ability to support the optimization of multiple criteria in the query process. This paper surveys the remarkable efforts made towards the extension of the aforementioned tools to overcome their limitations, respectively the explosion of the output result and the difficulty of query formulation. Moreover, we explore the application of these state-of-the-art techniques as preference-based query frameworks, proposing a comparison of their query personalization capabilities, the ability to control the output size and their flexibility with respect to the user input preferences.

show abstract

“…In reality, however, users typically have some pre-specified preferences over some dimensions of the data, while expecting good coverage over the other dimensions. Motivated by that need, we propose a novel scheme called ReDi, which aims to generate representative data that balance the tradeoff between regret minimization and diversity maximization [48]. ReDi is based on a hybrid objective function that combines both regret and diversity.…”

Section: Query Level Optimizationsmentioning

confidence: 99%

Scalable diversification for data exploration platforms

Khan¹

View full text Add to dashboard Cite

Today, data exploration platforms are widely used to assist users in locating interesting objects within large volumes of scientific and business data. In those platforms, users try to make sense of the underlying data space by iteratively posing numerous queries over large databases.Hence, data exploration platforms rely on methods for the extraction of representative data to provide users with a concise and meaningful representation of query results. That is, extracting a few tuples from a query result to provide quick insights in the potentially huge answer space.In the past few years, importance of diversification while extracting representative subsets of data has been greatly emphasized. It has been shown that diverse subsets provide more effective representation of the underlying data by minimizing redundancy and increasing coverage.Meanwhile, search results diversification adds additional cost to an already computationally expensive exploration process.In this PhD thesis, we have focused on the design, implementation and evaluation of scalable diversification algorithms and schemes for the data exploration platforms. Particularly, this research stipulates that extracting diverse representative results during data exploration requires addressing several challenges including: 1) scaling to big volumes of high dimensional data, 2) large number of users, and 3) enabling real time continuous exploration. To address those challenges we focus on two broad aspects: 1) Diversification of high dimensional large data sets and 2) Diversification of multiple user queries.The existing work conducts diversification in two steps: first compute all relevant query results, and then diversify the query results to select a small diverse subset. Similarly, all the dimensional attributes of all the data points in a query result are considered for diversification. Such a generic approach would be a performance bottleneck in high-dimensional large databases. To efficiently compute diverse subsets of query results exhibiting both high dimensionality and high-cardinality, we have proposed the Progressive Diversification Scheme. Our proposed scheme, utilizes the partial distance computations to reduce the amount of CPU and I/O incurred during query diversification. Moreover, to avoid the overhead of computing all relevant results first, we propose embedding diversification in query evaluation step by utilizing column-based data storage systems. In addition to computational cost of diversification methods, we have also considered the complexity of diversity objective function in high dimensional databases. Often, computing diverse solutions along all the dimensions is not a realistic approach. In fact, users may have some pre-specified preferences over some dimensions of the data, while expecting good coverage over the other dimensions. Motivated by that need, we propose a novel scheme, which aims to generate representative data that balance the tradeoff iii between regret minimization and diversity maximization. Our scheme...

show abstract

Diversifying with Few Regrets, But too Few to Mention

Cited by 8 publications

References 13 publications

Efficient Diversification for Recommending Aggregate Data Visualizations

Efficient Diversification for Recommending Aggregate Data Visualizations

Skyline Operators and Regret Minimization Techniques for Managing User Preferences in the Query Process

Scalable diversification for data exploration platforms

Contact Info

Product

Resources

About