The large design organisations which are carrying out complexes of design works within the limits of performance of projects “turnkey” in due course accumulate great volumes of the data which analysis by means of means OLAP allows to prepare operatively summary analytical reports for a management which are generated on a corporate portal of the company or a BI-portal.
Let's consider concept about a corporate portal
Before to pass to consideration of a role and a place of OLAP-technologies in a modern corporate portal, it is necessary to begin with an explanation: and what such a corporate portal? The corporate portal is “an access point” which provides external and internal users with the uniform, safe network user interface with the personified content. The content should be understood in the widest sense as everything that unites a portal in one user interface: appendices, the information and tools of joint activity. Internal either external operational or analytical hardware-software means concern appendices. The information represents the structured data, results of the analysis (reports, cubes, schedules, tables, etc.) The internal or external not structured content: documents, contents of digital stores, statistics of visiting of a site, etc. Tools of joint activity are web chats, network conferences, e-mail, services of instant messages, etc.
Let's consider the user interface of a portal and OLAP-technology
The recognised expert in the field of IS Mike Ferguson notices that a corporate portal and the intranet not same. Key distinction between a portal and the intranet – that portal technologies give the chance to personification that is not inherent in the intranet. Each user sees something distinct from that sees another depending on its function. The corporate portal leans against the specific technological products, which possibilities much more, than at the interface of the user of a network. They contain a full set of services.
One of aspects of a corporate portal – providing of access to the information from one point, irrespective of its type. According to researchers from Mjunstersky university, the body of the user interface consists of the list in which the relevant information of all types (video, the text, OLAP-reports, etc. is presented.), answering to a current context; its type can be visualised by means of labels or the explanatory text.
Scientists suggest to describe any of elements of a content the metadata of a corporate portal: the list of terms named a vector of terms (term vector). Such metadata gives the chance search irrespective of information type of a content. Thus, the vector of search also bearing in elements of structure of the metadata, displays a current interesting context and catalogues a content which vectors of terms answer a search vector.
|Drawing 1 – Contextual search match case with use of the metadata of a content of a corporate portal|
For visualisation of a current context the search vector is displayed in a line of a condition. At portal start, the search vector is initialized by terms which as it is supposed, are relevant for the user. Search function can be caused at any moment (drawing 1, operation 4). On the basis of structure of the metadata it is possible to make amendments to a current vector of search (drawing 1, operation 5). The call of record from the list of results starts the application visualising the corresponding content. The full vector of terms of a content is simultaneously displayed. As navigating operations (within a quantitative or qualitative content; drawing 1, operation 3) can lead to updating of a vector of terms, it is necessary to update it continuously that at any time gives the chance to conduct contextual search match case.
If it appears that any of OLAP-reports, does not correspond to a search vector, dynamic definition for the new report is supported. In this case terms of a vector of search reflect the technical metadata of the OLAP-report (help objects and factors) which can serve as initial definition of the new OLAP-report. The information which is not answering to definition of the report, is addressed by default.
Earlier mentioned scientists from Mjunstersky university notice that occurrence of network technologies and necessity of reception of the information from one source have raised the importance of portal technologies. That is why systems of storehouses of the data, and equally control systems of a content have been strengthened by functionality of portals and Internet interfaces. The publication of OLAP-reports with instructions of the unified index of an information resource (URL) and navigation support under the report in a network the Internet allows to carry out primary integration for construction of a uniform information field.
Speaking about a place of OLAP-technologies in a portal, interesting will result the point of view of experts from consulting company Object Systems Group. In their opinion, the sphere of application of OLAP-tools in a broad sense is “advanced” the business analysis (Business Intelligence). These tools can address to the data of various objects, correlate them, and also carry out information search (on subjects) in the scales which are not subject to the majority of users, in a case if it was necessary to do the same manually. For today, the majority of the companies already have many of the components necessary for introduction of OLAP-portals (drawing 2 see). Nobody begins to argue that possibility to take any business information, having asked a question “how it corresponds with in what I am engaged?” – the powerful tool of support of decision-making of which users can dream only. Really, using OLAP for generating of reports to a management of the organisation performing design works, it is possible to receive always “curtailed” the actual data.
|Drawing 2 – Typical components OLAP of portals|
Let's consider a BI-portal
Users of BI-tools (Business intelligence, Business analytics) are possibly familiar with BI-portals (the screenshot of a typical BI-portal is presented in drawing 3), which, according to mentioned above Mike Fergjussona, unlike a corporate portal more narrowly specialized in the sense that provide external and internal users with the uniform, safe, network interface for work with the personified integrated BI-tool (reports, cubes, tool panels, administrative panels and other tools).
|Drawing 3 – Screenshot BI of a portal|
In itself BI-portals quite often intend for providing of access to the information exclusively in connection with a tooling of the certain developer and not “are ground” on integration of all content is a problem of a corporate portal. So, it is represented useful to integrate these two kinds of portals, in particular, if in the organisation performing design works, portal BI-products are introduced, and is essential it is necessary to integrate all analytics from numerous and various B-appendices and tools into the uniform user interface to facilitate a life to each separately taken user.
The interesting point of view on the BI-portal organisation Warren Thornthwaite from university Kimballa. In its opinion if to recognise that the majority of users come on a BI-portal to find their interesting information, the considerable part of the main page of a portal should be issued in the form of the standard catalogue of the standardised reports and analytical notes categories. Basically, already it was possible to grope the best approach to the organisation of a BI-portal for service of key business processes of the enterprise. Their categorisation gives the chance to users to be defined quickly with a right choice by search. Each category contains detailed subcategories that allows not to waste time on long studying of the main page to find the interesting information.
Thornthwaite imparts experience and notices that header pages of BI-portals more often, than followed, are focused on the help information on history of creation and introduction of Storehouse of the data, a current condition of process of loading, and even persons of members of group of support. Certainly, it is very interesting, but, as a rule, to a BI-portal go not behind it. The BI-portal is the user interface to the Storehouse data. By its working out should consider inquiries of community of users first of all.
Let's consider movement from strategic to an operative or timely business analytics
In business intelligence delay reduction between occurring business event and undertaken action is important. This delay consists of three components: delays by preparation of the data (time which is necessary for preparation of the data for the analysis), an analytical delay (time which is necessary for reception of results of analytical operation and a delay in decision-making (time which is necessary for decision-making by the person who has received results of analytical operation). That the operative business analytics was effective, it is necessary to reduce time of these three delays practically to zero.
For reduction of a delay of the data it is created more and more virtual components business inteligence, including virtual Operational Data Store (ODS) and show-windows of the data with use of EII-technologies of the design organisations. If your operative data is in enough good form (the minimum integration and clearing of the data is necessary) for delay reduction in granting of the data can be used virtual ODS and show-windows of the data. However thus it is necessary to trace influence of the given environment on operative systems.
Such technologies are applied to reduction of an analytical delay as Business Activity Monitoring (BAM) or the operative information panels giving results of work to commercial users and sending alarm signals at excess of limiting values. Key Performance Indicators (KPI), the most important indicators, given each hour or more often, and the general operative results can be reflected by means of information panels or portals that allows operation personnel to familiarise with the basic occurring processes.
Though acceleration of process of gathering, the analysis and display of the operative data, undoubtedly, is useful for business, it is necessary to remember that not the business analytical information here should join. Many IT-experts do not spend careful check for that exact definition, what data should join in operative business analytical appendices. They do a catastrophic error, including in them so a lot of information as soon as it is possible, and reflecting all data “in a mode of real time” that leads to creation of difficult and unhandy business analytics.
It is required to study carefully a question, what analytics should be given in a mode of real time. It is necessary for you to understand requirements of business and then it becomes clear that only very small percent from total amount of the data is necessary for including in operative BI-appendices. The most part of the analytical data can become outdated on hours, days and months and at the same time to correspond to the basic processes.
Let's consider the tendency to more difficult analytics and visualisation of the data
In the beginning there were simple reports and inquiries, then time of the multidimensional analysis of the data has come. Though cubes and the scheme a star remain very popular, transition to more complex, deep analytics with use of the intellectual analysis of the data, statistical methods and the technologies supporting the analysis in style ad hoc now is observed. Probably, it is usual evolution of technology business intelligence, adapting to more developed business environment. Definitely, level of business knowledge is much higher now in comparison with what was 10 years ago that allows even to not qualified users to spend a complex analysis increasing possibilities of introduction and use of knowledge.
These technologies have raised level of knowledge of predicting analysts and ability to decision-making that allows to apply them in operative information streams. The companies are capable now to an operative or timely business analytics that gives the chance to the advanced workers of access to results of the given analytics and use of these results in a combination to the operative data in daily activity.
Also it is necessary to remember that the given possibilities do not arise in itself. They should enter into the general system business intelligence the organisation performing design works, and its technological infrastructure, in a return case the chaos will be established. The operative business analytics also demands full understanding of business processes or labour processes which it makes active. Without understanding of it, the experts introducing to a business analyst, will not know, as well as where with the maximum benefit to apply this valuable knowledge.
Thus, a BI-portal to construct not so it is simple, but it is the major link in a cost chain of storehouse of the data. Any word, heading, function or the reference included in a portal, should be connected with a subject content of storehouse of the data/business analyses. It is necessary to pay attention to portal check together with users, suggesting them to find certain reports and other information with the further critical analysis of the received results. In other words, it is necessary to be convinced that the system has no failures and weak links.
1. Use OLAP and other means BI in a corporate portal. – 2007 [http://www.iso.ru/journal/articles/501.html]
2. Imhoff C. Three tendencies in technology business intelligence. – 2008 [http://www.infology.ru/2008/09/08/578]
The author: Челябэнергопроект
Comments of experts of Челябэнергопроект: