可擴(kuò)展數(shù)據(jù)清理軟件平臺(tái)的研究

提出一種可擴(kuò)展的數(shù)據(jù)清理軟件平臺(tái)，該軟件平臺(tái)具有開放的規(guī)則庫和算法庫，規(guī)則庫用來存放清理規(guī)則，算法庫用來存放清理算法，算法庫中包含多種算法，并可對(duì)其擴(kuò)展；通過在規(guī)則庫中定義清理規(guī)則以及從算法庫中選擇合適的清理算法，可使該軟件平臺(tái)適用于不同的數(shù)據(jù)源，從而使其具有較強(qiáng)的通用性和適應(yīng)性；通過多種算法的清理，提高了數(shù)據(jù)清理的綜合效果。最后，通過實(shí)例驗(yàn)證了該平臺(tái)的效果及可行性。

An extensible data cleaning software platform is proposed, which has open rules library and algorithms library. Rules library is used to store rules and algorithms library is used to store algorithms. Algorithms library has many algorithms and can be extended. Through defining rules in rules library and choosing proper cleaning algorithms from algorithms library, the software platform can be used to various data sources, which makes it universal and adaptive. The synthetic result is improved through data cleaning with many algorithms. Finally, the effect and feasibility of this extensible data cleaning software platform is proved through an example.