Please use this identifier to cite or link to this item:
|Title:||Data visualisation of data patterns extracted from business processes|
|Abstract:||As more time passes, more data is saved inside businesses’ databases and in most cases this data becomes just archival data with no use at all. Since everyone started to talk about Business Intelligence during the last ten years, every business is finding it a necessity to implement BI in one’s organisation. Business Intelligence is made up of different processes and in this project two of these processes which are heavily linked together are studied and implemented. These are the extraction of data patterns from data sources i.e. Data Mining, and the visualisation of these extracted pattern, i.e. Data Visualisation. In this project four different data mining techniques are used which are Association Rules, Recommender Systems, Clustering and Time Series Analysis. All of them can be used to analyse this archival data and extract from it useful information which a business can really benefit from. For example by using association rule mining, one can extract those products which are normally bought together in a retail store. Afterwards this information can be used by marketing people to arrange shelf structure in order to increase sales of a lesser popular product. Normally a lot of time is spent on the extraction of patterns, however less time is spent on building a visual model on how these data patterns can be visualised. This project tries to emphasize this necessity by giving equal importance to both of them. After the extraction of such data patterns, different visualisation tools are used to better visualise the data patterns extracted by each technique. By giving this importance to data visualisation, the effort spent on the previous process will not be lost as in the majority of the cases when these data patterns are given to decision makers inside an organisation, they do not know how or spent a lot of time in trying to understand these extracted patterns.|
|Appears in Collections:||Dissertations - FacICT - 2015|
Files in This Item:
|3.09 MB||Adobe PDF||View/Open Request a copy|
Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.