Selection Contractors in E-Tendering Procurement Of Goods And Services Bureau Central Kalimantan Using Analysis Network Process

Selection a contractor in e-tendering procurement of goods and service is an routine activity at tenders of Central Kalimantan government. It task is doing the selection against the contractors of e-tendering that offer documents are uploaded on the website of the official Bureau of procurement of goods and services of Central Kalimantan. Selection a contractor is a complicated task because it must be transparently but the problem is the system still lacks proper judgment for not explained about score points for contractors who do not win. System use fall system and not give a thorough explanation about contractors who did not win the tender. This research aims to be able to resolve the problem using Analysis Network Process to select contractor to provide the weight of each of the criteria. The criteria are administrative, price, technical, and prove the qualifications must be completed by each participant contractor who register on e-tendering. The contributin this research serve as a guide for the Bureau of procurement of goods and services and select the contractor in tenders held so a clear assessment and is open to contractors who follow e-tendering held Bureau Procurement of goods and services in Central Kalimantan.


Introduction
Indonesia have many Imperial construction companies that is as much as over one hundred thousand construction companies in Indonesia. In Indonesia there are 128,570 contractors active nationally. Details of the number of small contractors as much as 108,626 or 85%. The number of big contractors only 2% or 2,433 contractors and contractor 17,511 medium or 14% [1].
Procurement Of Goods And Services Bureau in Central Kalimantan handle project tenders procurement of goods and construction working above 200 million rupiah. Projects under the 200 million rupiah is handled directly by the Agency through a direct appointmentrelated contractors. A tender on selection of Bureau of procurement of goods and services through several stages, namely the administrative file selection, selection of technical work, selection of price and selection of proof qualifications. The current bureau system this is a knockout. It is becoming one of the focus of the research is due to leave the system fall used to be an issue that is often on the talk about local contractors in Central Kalimantan since determination of the winners don't pay attention to the weight of each criterion points. In the selection of contractors is less appropriate if using knockout because of not considering the overall score points but only pay attention to the completeness of the file. Bureau of procurement of goods and services formed a team in each tender which held that is composed of a Chairman and two members. The team in charge of conducting the selection against the tender. Contractor selection with four stages, namely administration, technical work, price quotes, and Standard Procurement Documents based on qualifying goods/services the Government Electronically with E-Tendering (Regulation No. 4 of the year 2015) [2].
In the selection of contractors, the team has the task of deciding the winner of the tender on the basis of predetermined criteria. The team used the knockout in the selection of contractors in Central Kalimantan. In this study the author aims at contributing to the fall of the system into a system of weighting the scoring to be transparent on the selection of contractors. In this study use ANP (Analytic Network Process). The Analytic Network Process (ANP) i.e. selection with several stages, namely determining criteria, weighting, pairwise comparison, eigen vector, test the consistency of the ratio, then rank [3]. The Analytic Network Process (ANP) is a a multicriteria theory of measurement used to derive relative priority scales of absolute numbers from individual judgments (or from actual measurements relative to a normalized form) that also belong to a fundamental scale of absolute numbers [4].
This research aims to provide theoretical contribution about the ANP with the awarding of the weighting on each criterion because Bureau of procurement of goods and services (BPBJ) Central Kalimantan did not previously use the system weights but knockout so this study into consideration Committee working group decision in determining the winner of the tender for construction in Central Kalimantan.

Analysis Network Process
According to hosseini [5] The ANP methodology consists of three main steps, namely, pairwise comparisons and local priority vectors, weighted matrix, and super super matrix formation and transformation. ANP is used for the selection of the best strategy in supply chain risk issues.
Stages of the ANP, according to coulter et al [6] is a decision problem, using pairwise comparison, The relative weights or priorities of the elements, Relative local priorities of each element, Additional "files" of elements, Networks of files.
According to jeon et al. [7], the ANP are used in networking cooperation partners. Approach based on the analytic network process (ANP) to rank a set of potential candidates within a multivariate set of attributes systematically. The criteria used are the 25 criteria.

Pairwise Comparison
The use of pairwise comparisons for comparison matrix criteria is a lot of trouble if too many criteria. In addition, the ANP pairwise comparison procedure and subsequent analysis permits the integration of quantitative and qualitative /intangible/ tangible measures into a single overall score for ranking decision alternatives. Finally, the ANP is superior to other forms of multiattribute analysis [8] as well as to choice-based conjoint methods, in that it enables its user to consider interdependencies between and among factors.

Consistency Ratio
According to kabak et al [9] If the value of Consistency Ratio (CR) for the matrix is smaller than 10%, the DMs ' consistency for the evaluations within the matrix is acceptable. If the CR is equal to or greater than 10%, pairwise comparisons within the matrix must be evaluated again.

Methodology
Methodology used ANP with the following stages.

Weight of Criteria
The weighting of criteria based on the scale of the ANP saaty as follows:

Pairwise Comparison
Pairwise comparison is used in determine the decision to form the matrix so as to change the data into the ratio [11]. Paired comparison matrix mengkalikan i.e. himself against himself as below.   (1) On matrix consistency, practically , While not every variation of the matrix will bring changes to the value of . Deviation from n is a parameter of the Consistency Index (CI).

Description: CI= Consistency Index
= maks eigen value n = the number of elements that are further than did the calculation equation with the formula CR : (2) Description: CR = Consistency Ratio CI = Consistency Index RI = Random Index The ratio of the consistency or CR obtained using equation 2.2 The value of the RI for n = 4 is 0.89 can be seen in table 2.2 CR = CI/RI = 0/0,89 = 0 The above value is consistent because CR ≤ 0.1. If the value of CR > 0.1 then not consistent or unqualified and subsequent decision matrix should be repeated until the CR values consistently or qualify the consistency.
6. Determine value alternative -Value alternative The value of the specified range, the data range that is created into a scale of interests saaty, can be seen in the following table: Table 6. Range alternatif comparison

Range
Value of interest Very good-very good 1 Very good -good 3 Very good -enought 5 Very good -not good 7 good -enought 2 good -not good 4 enought -not good 2 Here is an example of the case selection in Central Kalimantan using method ANP. Pairwise comparison done between subsequent alternative against each of the criteria and looking for the value of the eigen vector.  From the matrix above, eigen vector can be determined, lambda consistensi index and maximum, and consistensi ratio. Eigen vector values obtained from the division between the first line against the number of values in the first column of the second row in Add and divide the amount of the value of the second column and so on up to the third column. The first column of the matrix sum of: 3 + 3 + 3 = 9 The second column of the matrix sum of: 3 + 3 + 3 = 9 The third column of the matrix sum of: 3 + 3 + 3 = 9 Next do the calculations to find the following eigen vector.

Conclusion
In this research it can be concluded that for the selection of contractors in Central Kalimantan with the method of Analysis Network Process is with the stages determines the criteria first, determine the value of the criteria, then it is doing Comparative Matrix pairs and then the results matrix consistency tests conducted trials to prove that the data used can be used or not, if the value consistency less than 0.1 then the data is consistent and can proceed next step determine the value of the eigen vector in each criteria. Alternative input that is done in the next 3 contractor PT X, Y, and Z values and are given more weight in accordance with the scale of fundamental values and calculations done saaty eigen vector on each alternative against the criteria. After all the stages already done then the next stage is done rank alternative. The highest rank results will be the contractors. Phases of Network Analysis Process used is to specify criteria based on qualitative data from the Bureau of procurement of goods and services.