دانلود رایگان ترجمه مقاله انبار داده ها و سیستم پردازش تحلیلی بر خط داده های خیلی بزرگ – ACM 2013

انبار داده ها و OLAP داده های بزرگ: چالشهای موجود و سمت و سوی مطالعات آتی

Data Warehousing and OLAP over Big Data: Current Challenges and Future Research Directions

خلاصه
در این مقاله ما به مسائل موجود و سمت و سوی مطالعاتی در زمینه ی انبار داده ها و OLAP داده های بزرگ میپردازیم. “داده های بزرگ” اصطلاحی است که اخیرا در این حوزه مطرح شده است. همچنین به افق های مطالعاتی پیش روی جدیدی در این حوزه اشاره خواهیم داشت و نتایج مثبتی که طی این نوع مطالعات حاصل خواهد شد را مطرح خواهیم کرد.

ABSTRACT

In this paper, we highlight open problems and actual research trends in the field of Data Warehousing and OLAP over Big Data, an emerging term in Data Warehousing and OLAP research. We also derive several novel research directions arising in this field, and put emphasis on possible contributions to be achieved by future research efforts.

۱٫ INTRODUCTION

A strong interest towards the term “Big Data” is arising in the literature actually (e.g., [12,17,18]). This term identifies specific kinds of data sets, manly of unstructured data, which populate the data layer of scientific computing applications (e.g., [21]). Data stored in the underlying layer of all these application scenarios have some specific characteristics in common, among which we recall [1]: (i) large-scale data, which refers to the size and the distribution of data repositories; (ii) scalability issues, which refers to the capabilities of applications running on large-scale, enormous data repositories (i.e., big data, for short) to scale over growing-in-size inputs rapidly; (iii) supporting advanced ExtractionTransformation-Loading (ETL) processes from low-level, raw data to somewhat structured information; (iv) designing and developing easy and interpretable analytics over big data repositories in order to derive intelligence and extract useful knowledge from them. Due to the intrinsic nature of Big Data application scenarios (e.g., [25]), it is natural to adopt Data Warehousing and OLAP methodologies [2] with the goal of collecting, extracting, transforming, loading, warehousing and OLAPing such kinds of datasets, by adding significant add-ons supporting analytics over Big Data (e.g., [11,1,21]), an emerging topic in Database and Data Warehousing research. Data Warehousing and OLAP are classical scientific fields which have been addressed since several decades by the Database and Data Warehousing research community. Symmetrically, the fundamental problem of computing OLAP data cubes has been contextualized in a wide family of types, ranging from classical relational data sets (e.g., [3]) to graph data sets (e.g., [4]), and from XML data (e.g., [5]) to novel social network data (e.g., [6]), and so forth. With the advent of the Big Data research context, it is natural to think of the problem of computing OLAP data cubes over Big Data as one of the top-interesting challenges in the research community, with also powerful technological achievements to be reached within the scope of real-life large-scale data-intensive applications and systems. Unfortunately, despite the clear convergence, state-of-the-art solutions are not capable to deal with computing OLAP data cubes over Big Data, mainly due to two intrinsic factors of Big Data repositories: (i) size, which becomes really explosive in such data sets; (ii) complexity (of multidimensional data models), which can be very high in such data sets (e.g., cardinality mappings, irregular hierarchies, dimensional attributes etc.). As a consequence, there emerge the forceful needs of designing novel models, techniques, algorithms and computational platforms for supporting the problem of computing OLAP data cubes over Big Data, which, indeed, literally represents an effective call to arms for next-generation Data Warehousing and OLAP research. Inspired by this main motivation, in this paper we highlight open problems and actual research trends in the field of Data Warehousing and OLAP over Big Data (Section 2), (2) derive several novel research directions arising in this field, and (3) put emphasis on possible contributions to be achieved by future research efforts.