Business Intelligence Issues - Essay Example

RUNNING HEAD: BUSINESS INTELLIGENCE APPROACHES.  Topic: Business Intelligence Approaches Name: Student No: Institution: Business Intelligence Approaches Introduction This assignment shall evaluate three approaches of application of Business Intelligence (BI). Business intelligence has a broad category of technologies and applications which are applicable in different sectors of the economy and are used in aiding of decision making. The applications collect data, analyze it, generate highly customizable reports and store the data and the information processed in an easy way for its retrieval. The information analyzed has historical factors, current operations and future operations of any business organization. Needs of decision making by those in position of decision making are increasing being more and invention of Business Intelligence technology and applications is handy as it enable them make high quality decisions which improves their operations due to its high degree of efficiency, accuracy and reliability which is required by any decision maker. For the purpose of this research it will analyze three business intelligent approaches which are Standard Statistical Methods for Quantitative Data that are applied by use of forecasting, predictive models, decision trees and neural nets. Semantic Analysis methods which uses Latent Semantic Analysis (LSI) of textual data and Geographic Information Systems for spatial data. Quantitative Data Analysis Quantitative data analysis is used in a wide range of application especially those which use numerical data that need to be analyzed by business intelligent applications. The process involves presentation of numerical data which need to be analyzed. Analysis being done on the data would be Inferential statistics or descriptive statistics. Descriptive statistics will include measurements of tendencies which would be done by calculation of mean, mode and median. It also measures variability by use of standard deviation and range. Such analysis enables the decision maker to make conclusions of the data which are mainly got from research or business operations. The other method which quantitative analysis business intelligent application does is Inferential Statistics which are got as outcome of statistical tests they help deductions to be made from the data which is being collected to test hypotheses set to relate findings with the sample or population. According to Schneider, (2005) there are different types of quantitative measures which are used by Business Intelligent applications they include Nominal or category which is done by enumeration of categories, ordinal done by use of ordered scales and lastly interval or quantitative also know as scale which is done by (SPSS) which measurement are done in intervals (Connolly, 2006). Statistical analysis has principles of finding the structure of data this is done by either adding Non-Structured and Structured or obtain data structure from a combination of non explained variance and explained variance. For the business intelligent system to be effective it is important to adhere to stages of statistical analysis which involves; collection of date and cleaning it to ensure the date to be processed is correct, secondly is gaining knowledge about the data which is being used so that one would decide on the appropriate application to use to analyze the data. The applications which are used to analyze the data are decision trees, decision tables and neural nets. They are used mainly for forecasting in occurrences of near future and use of predictive models which are able to generate future treads based on the past and current occurrences. The visualization method which is used by these business intelligent applications is for foreseeing the future. Most decision makers in a business environment are much worried of the future of the organizations and they always base their current decisions to shape the future of the organization. Therefore uses of business intelligent applications furnish the administrators with information which enable them evaluate their current operations and if not in line with their future goals they can reevaluate it. The approach is effective as it enables the decision makers to make timely decisions which are accurate. The arithmetic calculations and arguments which are done by the application are effective as they work out complex computations which if they were to be done manually would take much longer and they are not guaranteed of almost 100% accuracy. Semantic Analysis Methods (LSI, LSA) of textual data. Latent Semantic Analysis (LSA) is a thesis and methodology of extraction and representation of contextual-usage defining words by arithmetical computations useful to a large amount of transcript. According to Zhang& Mostafa, (2002) A technique of analyzing relationship linking a set of credential and terms they foster by production of concept sets allied to the credential and terms particular in vectorial semantics analysis. In the milieu of LSA its application to informational reclamation is also known as Latent Semantic Indexing (LSI). Latent semantic analysis represents the number of words found in a large machine readable context after the processing them. Neural net models are also relatively fostered by LSA though the thesis is based on singular value decomposition; a technique of mathematical medium decomposition closely fosters factors analysis applicable to transcript corpora imminent volume of relevant idiom practiced by people. These patents was introduced in the year 1988 by a group of analyst The term document matrix is used by LSA to define the occurrences of terms in contexts; sparse matrix is when rows correspond to terms and when columns correspond to contexts, an example of typical weighing of element of matrix is tf-idf (term frequency-inverse context frequency): the number of times the term is featured in each context have to be proportional with the element of matrix. Standard semantic models also fosters matrix, however it’s not necessarily explicitly addressed as matrix, mostly because mathematical properties of matrices are skipped. The use of LSA has significantly expanded in due years as disputes in scalability and recital have been overcome. And many firms have appreciated and embraced the use of text in semantic basis in modern information retrieval systems. The primary application methods are for automated context categorization and conception searching. LSA can be used in the following ways; text summary, for the discovery of information, classification of automated context, when matching papers of technical importance and reviewers grants, in system administration for filtering spam, when used by engineers to code software sources, assist the government in linking charts of individuals and organizations for automatic generation, helps customer management by managing online customer support, in education the system is used for scoring of essay and information visualization and determining authorship of the context and materials of literature. Increase in use of LSA for discovery of electronic context (eDiscovery) this has been helpful to corporation preparation for litigation. The ability to categorize, cluster and search large assortments of unstructured transcript on a conceptual basis is mandatory. As early as 2003 leading providers of concept-based searching have bee using LSA in applying eDiscovery. There are certain drawbacks when using LSA analysis this may include; difficulty in interpreting resulting dimensions, the result can only be justified mathematically but have no interpretable meaning in layman and in natural terms. Because LSA represent each word as a single summit in space it cannot capture multiple definition of word. LSA is able to incarcerate and signify significant mechanism of the lexical and passage definition evinced in opinion and patterns by human beings from the results of formal and informal conjunction (Tristan, 2003). The problem which is solved by this approach is for analysis of textual data which is done by use of computer programs. There are different programming languages which are used to design the program however the working concept are the same. The program utilizes textual data for its analysis this data mainly are character and numeric which the program would even determine their format such as date, currency or percentage. The analytical technique which is used is manly comparing the data a good example is the search utility which is in most programs and one would be able to relate to. When searching one key in a word or a phrase which need to be searched and when they instruct the computer to search it peruses through the content as it search for what it has been instructed. In its operations it utilizes argumentative operations which are more similar to the operations of human brain therefore it is able to made decisions like human brain. With a very high speed and accuracy hence being if benefit when they are used by decision maker who want to analyze textual data. It makes us of functions and operators when computing the data the functions being used would be If…then, If…Then…Else, If…Then…Else If…Else, Do…While, Loop…Until and operators such as greater than, less than, addition, subtraction, division and multiplications Geographic Information Systems for spatial data Geographic information system (GIS) is a system of data storage system that stores, captures, manages, and represents data which is linked to a certain location. In technical terms, GIS systems are inclusive to mapping software and it’s closely associated with land surveying, aerial photographing, geography and all GIS software compatibles. Though referred to GIS this system does not cover all tools associated with topology. This can also be referred as tools applicable and allows the user to create interactive queries, spatial information analyzing, maps and presentation of all these operation in generic sense of view. The science underlying GIS is known as geographic information science this is a course study providing application, systems and geographical concept to those taught in GIS certificate and degrees programs in many universities and learning institutions. In layman term, GIS is the combination of database and cartography technology. Familiarity of application services offered to the consumer like finding driving direction by the use GPS (global positioning system) programs on their hand device or via mobile phone. This is just an example of how consumer interacts and relates with GIS services (Tomlinson, 2008). The GIS system of data collection can be dated back to the days of early man, evidence is shown and scholars attributes, that this is the oldest kind of data collection system, that has withstood time and is advancing to the future still yielding positive results. It’s estimated that between 20,500 to 15,000 years back in these era hunters were drawing pictures of animals they hunted in caves, along were tallies and track lines that analyst thought that depicted the migration routes as discoveries have it in records. A simple system of GIS was so created and this early man trend is associated with GIS in the images drawn were associated with attribute information. This is from discoveries in France in Lascaux cave. The first true GIS system to fully implemented and working was developed in the year 1962 in Ottawa Ontario Canada by the rural development and federal department of forestry, developed by one roger Tomlinson and named it Canada Geographical information system (CGIS) GIS is easily accessible because the use of software to attain these services is booming and free sites are readily accessed like the Google earth and more software’s firm are offering these services as incentives. GIS is used all over our community in learning institutions and in government organization. The data transferred here is of great importance and of every day need of our daily life. The systems are used to solve problems in geographical data the data which are used by the system are vectors which are a precise point in a geographical region this could be a specific point such as a bus terminal, a spring/river source or even a well it is very specific. The other form of data is pictorial by use of aerial photograph for a geographical region. Use of line is also popular in specifying, roads, rivers or rail line this are for structures on the earth surface which are liners. The other common data used is polygons which specify a common region such a city boundary, water masses or an enclosed region such as an institution (Peters, 2008). They are important in enabling geographers to analyze their data for environmental decision making based on the date obtained which the application captures the date, analyze process information and store it. References Connolly, (2006) Quantitative Data Analysis in Education: A Critical Introduction Using SPSS. New York: Routledge Peters, D. (2008). Building a GIS: System Architecture Design Strategies for Managers. Boston: ESRI Press Schneider, D. (2005). Quantitative Data Analysis. Retrieved on from December 09, 2009 from http://tecfa.unige.ch/guides/methodo/edu-tech/slides/analysis-quant.pdf Tomlinson, R (2008). GIS Hall of Fame. Retrieved on from December 09, 2009 from http://www.urisa.org/node/395 Tristan, M. (2003). Essay Assessment with Latent Semantic Analysis. Journal of Educational Computing Research. 29(4), 495-512 Zhang, J & Mostafa, J. (2002). Information Retrieval by Semantic Analysis and Visualization of the Concept Space. D-Lib Magazine. 8(10), 24-57 Read More