Research On The KM Evaluation Index And Support System In Software Process Improvement

Software Process Improvement (SPI) and Knowledge Management (KM) are respec-tively subject to software engineering subject and management science subject. In recentyears, an intense trend of using KM theory to guide and normalize SPI becomes more andmore clearly. Software organizations and Software Process Improvement are all knowledge-intensive. Knowledge Management can transmit the most appropriate knowledge to the mostappropriate people at the most appropriate time. Effective KM plays an important role inSPI, which can promote the holistic productivity and sustainable development ability of SPIorganizations.The SPI/KM evaluation index mainly includes SPI/KM maturity evaluation, SPI/KMprocess execution quality metrics and SPI/KM performance evaluation.This paper system-atically analyzes current domestic and international researches on the SPI/KM evaluationindex. Based on an in-depth study on SPI/KM evaluation index, this paper mainly focusedon the following issues:SPI/KM maturity evaluation methods and models, SPI/KM processexecution quality metrics and SPI/KM performance evaluation methods, KM-based SPI sup-port system iManager.Firstly, this paper puts forward some fundamental models which can be used in SPI/KMmaturity evaluation, such as a qualitative model KMCE and a quantitative model Weighted-3DCube. Based on quality management and process project, these models are the same withthe SPI organizations.Secondly, this paper studies on the qualitative and quantitative description methodsabout the SPI/KM process execution quality and the SPI/KM execution inconsistence de-tection. Based on the Weighted-3DCube model, this paper proposes two metrics for SPI/KM process execution quality evaluation, such as execution quality (SPIKM Quality), executionlimitation (SPIKM Limitation). The event tree based inconsistence detection methods putin this paper can be used to detect the inconsistence between process model and processexecution.Thirdly, after analyzing the existing performance evaluation methods , this paper putsforward a SPI organization oriented performance evaluation system, which consists of anindex system and an evaluation procedure. The evaluation system mainly focuses on thefollowing four factors, such as organizational intelligence, operation ?ow, customer serviceand employee development; contains 4 first level indexes and 18 second level indexes. Basedon the AHP (Analysis of Hierarchy Process), this paper defines a series of KM performanceevaluation process including index set and relationship definition, judgement set definition,index weight definition, subjection definition, integrated evaluation,etc..Finally, this paper brie?y introduces a KM-based SPI support system called iManager,which is partly designed by the author of this paper. This system is a project managementand process improvement support platform. iManager is a reduced CMM system, and isintegrated with the KM models and methods. It can be used as a knowledge share andcollaboration platform. The models and methods presented in this paper have been validatedin this widely used system.