A collaborative filtering approach for recommending OLAP sessions

Abstract

International audienceWhile OLAP has a key role in supporting effective exploration of multidimensional cubes, the huge number of aggregations and selections that can be operated on data may make the user experience disorientating. To address this issue, in the paper we propose a recommendation approach stemming from collaborative filtering. We claim that the whole sequence of queries belonging to an OLAP session is valuable because it gives the user a compound and synergic view of data; for this reason, our goal is not to recommend single OLAP queries but OLAP sessions. Like other collaborative approaches, ours features three phases: (i) search the log for sessions that bear some similarity with the one currently being issued by the user; (ii) extract the most relevant subsessions; and (iii) adapt the top-ranked subsession to the current user's session. However, it is the first that treats sessions as first-class citizens, using new techniques for comparing sessions, finding meaningful recommendation candidates, and adapting them to the current session. After describing our approach, we discuss the results of a large set of effectiveness and efficiency tests based on different measures of recommendation quality

Benchmarking Exploratory OLAP

“…We created an experimental setup to compare the following approaches: CineCube [13] and Falseto [1]. CineCube is a multifaceted approach focusing on building a user-friendly sequence of explanations for the analysts.…”

“…The user will then explore the surrounding region of the cube by means of OLAP operators such as roll-up (at the Europe level for example), drill-down (at the month level for example) and slices (for other products) to find evidences that may explain and corroborate the first fact. The user might even get some support from a system that automatically proposes next moves in the analysis [1,13]. We consider that the surrounding region of the first interesting fact corresponds to a neighborhood that has to be covered to ensure the exploration task success.…”

“…First we evaluate the benchmark protocol and metrics based on synthetic SUTs whose behavior is well known. Second, we show that it is possible Input Category DB instance Query log Current query Output [5] Automatic exploration Tuples [1] Automatic exploration Sequence of queries [12] Automatic exploration Queries [3] Automatic exploration Queries [13] Visual optimization Queries Automatic exploration result highlighting [14] Visual optimization Query [29] Data prefetching Tuples [28,30] Data prefetching Tuples [18] Data prefetching Sequence of queries [27] Data prefetching Queries Table 1. Interactive cube exploration techniques signatures and meaningful to compare two state-of-the-art SUT from IDE literature [1,13] with our benchmark.…”

SM4MQ: A Semantic Model for Multidimensional Queries

“…The evaluation showed that SM4MQ supports automation of transformation tasks that would otherwise require significant manual efforts (e.g., at least 100 times more just for SPARQL queries). In our future work, we plan to work on the exploitation side of the queries, e.g., develop richer transformations to support advanced user support techniques such as [4]. We also plan to apply our method to other use cases and support analytical feature definition via high-level GUIs.…”

“…MDM provides a GUI that enables a user to generate MD queries. It automatically generates SM4MQ query metadata as well as OLAP queries in Cube Query Language (CQL) [10] used by QB4OLAP explorer 4 for querying of QB4OLAP data cubes. The SM4MQ queries are stored in the SPARQL endpoint.…”