ISSN 2278-3083 International Journal of Science and Applied Information Technology (IJSAIT), Vol. 3 , No.3, Pages : 32 - 35 (2014) Special Issue of ICCET 2014 - Held during July 07, 2014 in Hotel Sandesh The Prince, Mysore, India Abstract : In this paper I presented a detailed summarization of the main features of each method regarding the criteria introduced in, which provides a common framework to compare and discuss methods surveyed. Methods surveyed are distributed in these tables according to the chronological order. On the other hand, some new approaches focus on considering alternative scenarios than relational sources. I presented the most relevant methods introduced in the literature and a detailed comparison showing. All in all, I discussed the current scenario of multidimensional modeling by carrying out a survey of multidimensional design methods the main features of each approach. Key words : Data warehousing, dimensional design methods, framework, Factual data, Dimensional Data. INTRODUCTION The data warehouse is a huge repository of data that does not tell us much by itself; like in the operational databases, we need auxiliary tools to query and analyze data stored. Without the appropriate exploitation tools, we will not be able to extract valuable knowledge of the organization from the data warehouse, and the whole system will fail in its aim of providing information for giving support to decision making. OLAP (On-line Analytical Processing) tools were introduced to ease information analysis and navigation all through the data warehouse in order to extract relevant knowledge of the organization. This term was coined by E.F. Codd in (Codd,1993), but it was more precisely defined by means of the FASMI test that stands for fast analysis of shared business information from a multidimensional point of view. This last feature is the most important one since OLAP tools are conceived to exploit the data warehouse for analysis tasks based on multidimensionality. The multidimensional conceptual view of data is distinguished by the fact / dimension dichotomy, and it is characterized by representing data as if placed in an n-dimensional space, allowing us to easily understand and analyze data in terms of facts (the subjects of analysis) and dimensions showing the different points of view where a subject can be analyzed from. One fact and several dimensions to analyze it produce what is known as data cube. Multidimensionality provides a friendly, easy-to understand and intuitive visualization of data for non-expert end-users. These characteristics are desirable since OLAP tools are aimed to enable analysts, managers, executives, and in general those people involved in decision making, to gain insight into data through fast queries and analytical tasks, allowing them to make better decisions [1]. Developing a data warehousing system is never an easy job, and raises up some interesting challenges. One of these challenges focus on modeling multidimensionality. Nowadays, despite we till lack a standard multidimensional model, it is widely assumed that the data warehouse design must follow the multidimensional paradigm and it must be derived from the data sources, since a data warehouse is the result of homogenizing and integrating relevant data of the organization in a single and detailed view. UNIT- I GENERAL ASPECTS The general criteria are summarized into nine different items: • Paradigm: According to (Winter & Strauch,2003), multidimensional modeling methods may be classified as supply-driven, demand-driven or hybrid approaches. The reader may found a slightly different classification in (List et al., 2002). Furthermore, we distinguish between sequential and interleaved hybrid approaches (depending if their supply-driven and demand-driven approaches are performed either sequentially or simultaneously or sequentially) [3]. • Application: Most methods are semi-automatic. Thus, some stages of these methods must be performed manually by an expert (normally those stages aimed to identify factual data) and some others may be performed automatically (normally those aimed to identify dimensional data). In general, only a few methods fully automate the whole process. On the contrary, some others present a detailed step-by-step guide that is assumed to be manually carried out by an expert. • Pre-process: Some methods demand to adapt the input data into a specific format that facilitates their work. For example, these processes may ask to enrich a conceptual model with additional semantics or perform data mining over data instances to discover hidden relationships. • Input abstraction level: Most methods (mainly those automatable) work with inputs expressed at the logical level (e.g., relational schemas), whereas some others work with inputs at the conceptual level (e.g., from conceptual formalizations such as ER diagrams or from requirements in natural language). • Output abstraction level: Several methods choose to directly generate a star or snowflake schema, whereas some others produce multidimensional conceptual schemas. Although many approaches argue that the data warehouse method should span the three abstraction levels, only a few of them produce the conceptual, logical and physical schema of the data warehouse. • Data sources: There are three items summarizing the main features about how data sources are considered in the method. ◦ Type of data sources: The input abstraction item informs about the abstraction level of the input, whereas this item specifies the kind of technology of the data sources supported by the method. For example, if the method works at the conceptual level it may work from UML, ER conceptual Comparison Of Design Methods Of Data Warehousing Tiruveedula GopiKrishna Sirt University, Hoon,Libya, gktiruveedula@gmail.com