Research On Decision Support System For A Construction Industry ERP Public Platform

The rise of the construction industry ERP software provides a great convenience for enterprise management, vendor maintenance, and selection of suppliers. ERP platform stores all the goods on offer by the manufacturers and suppliers in the database. Contractors can easily allocate the good they require and ask for bidding. However, with the increase in operating time, and the number of vendors increasing rapidly on the system, how to help engineering companies make suitable supplier selection from a large database of candidates is a question worth considering. Decision Support System provides a number of methods to help engineering companies converting their preference data to a form that computer can process with respect to each vendor attributes. With the advent of social network algorithm, we can select enough suppliers according to order requirements, so that it not only improving the efficiency of algorithm, but also to ensure the effectiveness of the screening results.This thesis proposes a decision support system based on social network algorithm to solve this problem. This research studies the major sub-contractors’ preferences in property acquisition, database data preprocessing, preprocessing data and engineering companies interested in screening data and adjust the intermediate and final data generation. This paper has accomplished the following tasks:1. AHP is used to get the attribute preferences suggested by the engineering managers. Then save the results to server.2. Depending on the ERP system, server periodically cluster the existing suppliers and vendors in accordance with the social network algorithm clustering algorithm, and the effects of various parameters on the clustering results.3. When Contractors need tender to select suppliers, the interest in the properties and clustering results can be acquired to generate recommendation result quickly, contractors can modify the intermediate results. Finally, the results will be processed by the mathematical model to generate the final screening, then show the final results.