Business Intelligence - Essay Example

Running Head: BUSINESS INTELLIGENCE Business Intelligence Name Institution Date Business Intelligence Introduction There are different forms of business intelligence approaches, namely, Geographical Information System for spatial data which takes, keeps, analyses, handles and presents information related to location. In particular, a GIS is a system that consists of mapping software and its application to remote sensing, surveying of lands, aerial photographing, geography and many others. Therefore, GIS refers to any information system that combines, keeps, edits, analyzes, distributes, and presents geographic information. GIS applications are tools that permit users to develop user developed searches, scrutinize spatial data, edit information, maps and present the final results. Conversely, semantic analysis methods involve a theory and a process of taking out and in lieu of the contextual-usage significance of words through numerical computations functional to a big quantity of text. A standard quantitative method for statistical data represents discrete analogs for the analysis and computing different types of information. This includes even employment of statistical data approaches to even preoperative and intra operative factors for the patients going through an injurious control surgery (Giddens, 1990). Case Study 1 A financial services organization had contracted two external dealers to create and convey a new performance reporting invention for their financial correspondents. However, these two dealers were having severe quality problems. The main challenge was that one dealer from Texas was supposed to merge a two month individual investor performance information into a monthly summary and convey the information to an external printer. The printer at Massachusetts was supposed to receive and convert the merged electronic information into monthly performance reports. These reports were meant to compare the performance of every investor against developed financial benchmarks. Consequently, the dealer would print and mail numerous investor performance reports to every customer’s financial correspondents. Conversely, information was not combined accurately, there were grammatical error and charts were not appearing accurately and individual reports were being printed and sent to the erroneous correspondent (Dijk & Kintsch, 1983). Therefore, a business intelligence approach was used to come up with an end to end quality assurance procedure that would guarantee the creation of exact individual performance reports and that the right dealer report would be mailed to the suitable correspondent. Business Intelligence made use of process flow chart which a standard quantitative method for statistical data with the aim of documenting each dealer’s “as-is” process. In this method, one embodiment of the invention that is available provides a system that implements a resource-usage strategy in a compute farm. The compute farm's resources can be characterized as a group of diverse kinds of job slots whose main intention is to allocate them to dissimilar types of jobs. For instance, the situation slots can be classified into light interactive situation slots, grave interactive situation slots, group job slots, excess job slots and many other classifications. The system can make use a lining system to run the set of job slots. Throughout the operation, the system can get custom regulations which consist of resource-usage regulations and corrective-action regulations. Regulations used in resource can identify circumstances in which the resource-usage guiding principle is desecrated, and corrective-action regulations can stipulate situations upon which a corrective measures needs to be carried out. Next, the system can get resource-usage information which consists of job evaluating information and process monitoring information. Job monitoring data can illustrate resource usage of examined jobs, and process monitoring information can illustrate resource usage of examined processes (Dijk & Kintsch, 1983). After this, the system can then establish a resource-usage desecration through application of the resource-usage regulations to the resource-usage data. After that, the system can store up the resource-usage desecration in a violation database. Still, the system can then establish a corrective action through application of corrective-action regulations to a sequence of violations kept in the violation database. Subsequently, the system can carry out the corrective action, thus implementing the resource-usage rule in the compute farm. For every examined job, the job examined information can stipulate the period the examined job has been implementing, and the period the information that has been examined has been not in use. Additionally, for every examined process, the process monitoring information can indicate the quantity of processor time that the examined process used, and the quantity of memory that the examined process used. The process monitoring information can also designate if a user executed a straight login on a compute farm node, together with the procedures that the user generated throughout the login session (Dijk & Kintsch, 1983). Basically, a corrective action can be any action that is intended to correct the resource-usage rule violation. In particular, the system is normally useful in sending a notification to the users who were the cause of the sequence of desecrations which generated the corrective action. Moreover, the system can easily send a notice/notification to the user's manager and/or to a system administrator. To add on that, the system can end the offending job or process that brought about the series of violations which generated the corrective action. Normally, the flow chart indicates a procedure for implementing a resource-usage policy in harmony with a personification of the invention that is current. The method can start through getting etiquette policy which consists of resource-usage regulations which denote circumstances in which the resource-usage policy is desecrated, and corrective-action rules which identify circumstances in which a corrective action requires to be performed. Every job or procedure characteristically makes use of a range of resources, like, processor time, memory and others. A resource-usage regulation can identify the resource-usage models that are measured to be violations. Still, a resource-usage regulation can stipulate the resource-usage outlines that are compliant (Giddens, 1990). Application of corrective-action regulations to a sequence of violations allows the system to better choose a corrective action. In particular, the violation database can maintain track of the violations that certain job or process has done previously. This enables the system to characterize intricate corrective-action regulations the basis being “status” of the violation. Ideal corrective actions are; sending notifications to users who made the series of violations which generated the corrective action to occur. A notification can also be sent to the user's manager and to a system supervisor of the compute farm. Incase notification sending cannot correct the violation; the system can end the incorrect job or the incorrect process which led to the sequence of violations which generated the corrective action. Incase the system ends the job, the system can inform a user the job was terminated and give an explanation as to why the job was ended. After the process flow chart was developed, the errors were identified. As the faults were denoted, the procedures were redesigned to stop similar errors from occurring again. The kinds of errors and the quantity of every error was trailed and communicated to the two dealers daily for them to evaluate the real causes of the quality problems and improve their processes. The result was that quality improved severely over the one year development phase because of the standard open communication and the aspiration for a win-win business relationship (Giddens, 1990). Case study 2 A health organization wanted to build capacity to deliver AML training after a costly and hard crush by an external training dealer. In the period of the crush, a major customer-facing system was because of insufficient testing and capability faults on worker’s PCs. The main challenge was that the internal learning management system within the organization never permitted any automatic identification of apprentices; neither did it offer vigorous reporting and tracking potentials. Business Intelligence approach was implemented to come to assist the organization in evaluation of its options developing and delivering an original AML training system in five months. The initial investigation established that it first had to establish the originality of organization’s current learner management system (LMS). The method which the organization’s experts perceived suitable to use was Latent Semantic Analysis to analyze the originality of information in current learner management system (LMS). To carry out the LSA analysis, the texts in learner management system (LMS) were run in an SVD scaling to produce a semantic space on the title of the system. The 10 texts the system’s context (6097 words), on the learner management system (LMS)’s context (~4800 words) and excerpts from two known computer system’s context (~17000 words) were incorporated in the scaling. Since the semantic space developed by LSA depends on having numerous illustrations of the co-occurrences of words, the adding up of other textual materials assisted in availing the LSA analysis with extra examples of learner management system (LMS)’s context interrelated words to facilitate in characterizing the semantics of the domain. The LSA investigation end result was a 100 dimensional space consisting of 607 text units by 4829 distinctive words. To establish the efficiency of LSA's forecast, regarding the source of LMS’s content, the forecast was evaluated against forecasts prepared by two individual raters. The raters, who were extremely knowledgeable on several LMS systems, went through the system’s context separately and for every sentence within the context, identified the texts which were likely to have originated from a familiar source. Finally, the agreement between the two human rates was that the LMS’s content was likely to be original. Moreover, this method can also be useful in analyzing the quality of the system. The initial step would be to assess the quantity of semantic overlie with the presumed original texts. Two measures of the value of the essays were computed through the use LSA. The first scrutinized the quantity of semantic overlap of the essays by the original texts. Every sentence in the system’s content was contrasted against every sentence in the presumed original texts, and a score was given base being the cosine involving the system content sentence and the closest sentence within the presumed original texts. As a result, there was a sentence that was unerringly identical as a sentence in the original text, content was to be rated 1.0, whereas a sentence without semantic overlap with anything in the original texts got a cosine of 0.0. The system’s content was rated and the end outcome was very low and therefore this system is likely to have used contexts that were original. The fundamental idea is that the entirety of information regarding the entire word contexts upon which a given word does and does not make appearance presents an array of mutual limitations that basically determines the correspondence of significance of words and set of words to each one of them. Eventually, AML training system was rolled within the five months since the LMS’s content were found to be original and hence similar contents were used in developing this other system. Under BIS leadership, this team developed and rolled out an automated application in less than four months, resulting in remarkable progress in training content, delivery, and in staff input and buy-in. The trailing and acceleration methods employed in addition to Latent Semantic Analysis led in 93% achievement in an 8 week time, a 57% decrease in general timeline, with least escalation of "incompletes" to division leaders (Press Rains, & Latham, 1993). Case study 3 A certain Japanese company was granted a contract by the United States government to carry out a major road expansion. The Japanese company needed to develop a process of evaluating the land where the road was to be constructed. The major challenge was that the road is in a remote location and hence through assessing of the land was necessary and this made matters very hard. Business Intelligence approach was implemented to develop a Geographical information system for spatial information to allow the development to find a common ground for offering a framework for the analysis of the geographical land where the land is located. The company used the Geographical information system for spatial information, working with cross functional professionals in geographical field to assess the land. The result was a clear illustration of land geography whereby all elevations, tiny or big within the land were identified. Consequently, the company was able to complete the project since after doing the geographical analysis, there was significant information regarding the land in the GIS database which was of much help during road constriction (Bauman, & May 2001). The area was directly taken an aerial photograph. Consequently, soft copy work stations are utilized in digitizing features from stereo duos of digital photos. These systems permit information to be taken in two and three aspects, with heights measured unswervingly from a stereo pair by means of principles of photogrammetry. At present, analog aerial photos are scanned before being entered into a soft copy system, but as high technology digital cameras are very useful since there is no need of scanning. . In order to store the information in a digital format, the information can be printed on a paper or even on film maps. Afterwards, these can be scanned or digitized to give out digital information. A digitizer gives out vector information as an operator tracks positions, lines and polygon margins from a map (Becker, 1986). The digitalized information was then stored in GIS in the raster form. Thus the information was represented in grids. Since pixel is the tiniest single unit of an image, a grouping of the pixels leads to development of the image, different from the scalable vector models. This raster information type then indicated the idea of reality. In this case, the raster indicated a presence of an elevation within the land. The raster information was then stored in a JPEG form. This information was in GIS database which had layers of information, thus in this case there was a layer representing the evenness of the land whereby there was an elevation. However, this approach was not much successful since a lot of information was expected. This could have been attributed to the fact that raster information in most cases lacks the traits of features it displays and still raster does not allow more analysis potential more so for networks like road and rails (Becker, 1986). Conclusion In any business intelligence system, integration of more then one approach of BI is fundamental yet misleadingly challenging. Companies, businesses and organizations should make use of BI to support the main decision making processes since this can be an amazingly high-profile undertaking. However, integrating information from different sources across an organization has exclusive challenges. Fortunately, best practices and systems for BI-information integration are swiftly developing. Currently, organizations have countless very useful in BI approaches. There are several new processes and tools on the market, all with their own advantages and disadvantages and some are more efficient than others, depending on the organization's necessities and finances. Moreover, some of the business intelligence approaches can even be useful in biomedical matrices as well as in analyzing some conditions. For instance, these statistical data analysis can be used in monitoring patients with high mortality rate. References Becker, H. (1986). Writing for Social Scientists. Chicago: University of Chicago Press. Bauman, Z. & May, T. (2001). Thinking Sociologically. Oxford: Blackwell. Dijk, A., & Kintsch, W. (1983). Strategies of Discourse Comprehension. New York: Academic Press. Giddens, A.  (1990). The Consequences of Modernity. Stanford, CA: Stanford University Press Rains, L., & Latham, P., (1993). GIS modeling of land use and associated no point source. Hillsdale, NJ: Lawrence Erlbaum Associates. Read More