Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index Decision Support System for Poverty Social Assistance using SMART and AHP Haniel Dwi Septhian*. Aji Supriyanto Faculty of Information Technology and Industry. Information Technology. Stikubank University. Semarang. Indonesia Email: 1,*thianlondo@gmail. com, 2ajisup@edu. Correspondence Author Email: Thianlondo@gmail. Abstract Poverty is a multidimensional problem that requires prompt and appropriate handling to maintain a dignified human life. In Manyaran Sub-district. Semarang City, the distribution of social assistance often faces obstacles due to limited human resources and a manual selection process for recipients. Therefore, a Decision Support System (DSS) is needed to assist the selection process in a more objective and efficient manner. This study aims to develop a DSS for determining social assistance recipients in Manyaran Sub-district by combining the Simple Multi-Attribute Rating Technique (SMART) and Analytical Hierarchy Process (AHP) methods. AHP is utilized to determine the weight of each criterion, while SMART is used to calculate the final score of each recipient candidate. The combination of SMART and AHP allows for both expert-based prioritization and quantitative evaluation, enhancing transparency and consistency in the selection process. The research was conducted through stages of problem analysis, data collection, literature review, system design, and report writing. The results show that among the ten analyzed candidates, the individual coded P06 achieved the highest final score of 0. The top five candidates with the highest scores were declared eligible to receive social assistance, while the others were declared ineligible. The application of the SMART and AHP methods in this DSS effectively improves the accuracy, objectivity, and efficiency of the selection process for social assistance recipients in Manyaran Sub-district. Keywords: Poverty. Decision Support System. SMART. AHP. Social Assistance INTRODUCTION Poverty is a multidimensional and multisectoral issue with various characteristics, which constitutes an urgent condition that must be addressed immediately in order to maintain and develop a dignified human life . Therefore, efforts to alleviate poverty must be carried out synergistically between the government, the community, and the business sector . One concrete step to address this issue in Manyaran Subdistrict. Semarang City, is through the establishment of a regional regulation on the provision of social assistance to poor communities. As a classical social phenomenon, poverty has long been embedded in community life. Although various efforts have been made, the belief that poverty cannot be completely eradicated only reduced and its suffering minimized still holds true . In practice, the large number of proposed aid recipients makes it difficult for village authorities to select those truly eligible for assistance . To ensure that social assistance is distributed accurately and that the welfare of village communities is achieved evenly, it is necessary to utilize current technology to support this process . Manyaran Subdistrict is one of the administrative areas located in West Semarang District. Semarang City. Central Java Province. Located at Jalan Simongan No. 200, this subdistrict covers an area of approximately 150 hectares consisting of residential yards and buildings, dry fields and gardens, sports fields, recreational parks, and cemeteries. Manyaran Subdistrict has 11 neighborhood units (RW) and 99 community units (RT), with a government structure consisting of six civil servants and twelve personnel in functional group positions . Various administrative services are available at the subdistrict office, such as the issuance of Family Cards (KK) and Identity Cards (KTP). The number of registered poor families in Manyaran Subdistrict reaches 393 KK, while the available personnel is very limited. This imbalance is one of the causes of delays in the distribution process of social assistance to the community. In addition, the selection process of eligible recipients is still carried out manually, which not only takes time but is also prone to errors and fraud. To address this issue, a system that can provide objective and efficient decisions is needed. One solution that can be implemented is to develop a Decision Support System (DSS) to determine recipients of social assistance. This system aims to prevent errors in aid distribution, considering that there are still beneficiaries in the field who do not meet the poverty criteria. With this system, it is expected that the aid will be received by those who truly need it. In building the decision support system, the Simple Multi-Attribute Rating Technique (SMART) and the Analytical Hierarchy Process (AHP) methods are used to enhance the selection process. SMART (Simple Multi Attribute Rating Techniqu. is a multi-criteria decision-making method developed by Edward in 1977 . This decision-making technique is based on the theory that each alternative consists of several criteria that have values, and each criterion has a weight representing its importance compared to other criteria . Meanwhile, the Analytic Hierarchy Process (AHP) method is a problem-solving method that evaluates alternatives against a set of attributes or criteria, where each attribute is independent of the others . Several previous studies have also examined the use of SMART and AHP methods in the development of decision support systems. Malisa Huzaifa and Evi Refianti . built a DSS for recipients of Village Fund Direct Cash Assistance using the SMART method with a success rate of 65. However, the study only used the SMART method without considering the weighting of criteria through other methods . Meanwhile. Bobby Ginting and Fricles Sianturi . used the AHP method to determine recipients of aid for underprivileged families with results of CR < 0. 1, but they also did not integrate other methods into their ranking process . Other studies such as Pratama . used the SMART method to determine the eligibility of cooperative aid recipients, but only based on four criteria with a limited scope . On the other hand. Yulrio Brianorman . Copyright A 2025 Authors. Page 42 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index combined AHP and SMART methods in a DSS for determining promotional regions, but with a different research focus, namely on regional promotion, not social assistance . Similarly, the study by Meliana Sabet Tambunan . used the AHP-SMART combination to select the best teachers, but the research object and focus differed from the issue of poverty . Based on these previous studies, this research develops a decision support system for determining poverty-related social assistance recipients in Manyaran Subdistrict by combining the SMART and AHP methods. This combination aims to increase the objectivity of criteria weighting and produce a more accurate ranking of potential aid recipients. The results of the decision support system will include calculations and rankings of prospective recipients, which will then serve as recommendations for village officials in making decisions. Thus, the service of village officials to the community can run more optimally and fairly. RESEARCH METHODOLOGY 1 Research Framework In this study, the researcher employed a systematically structured research method to accurately achieve the research The research stages are described in the following flow diagram: Problem Analysis: The first stage carried out by the researcher was analyzing the problems occurring in Manyaran Subdistrict. Semarang, related to the provision of social assistance for impoverished communities. Through this analysis, the researcher identified the need for a Decision Support System (DSS) that could facilitate a more objective, fair, and measurable selection process for aid recipients. Data Collection: The data collection in this study was conducted through two primary methods, namely observation and interviews. The observation method involved directly monitoring and recording the actual conditions, activities, and processes occurring in the field, specifically within the Manyaran Subdistrict. This allowed the researcher to gather contextual information regarding the socio-economic circumstances of local residents. The interview method was carried out with Mrs. Nina Rizqiana Nugrahaeni. SE, who serves as the Head of the Economic and Social Welfare Division at the Manyaran Subdistrict Office. As a subject matter expert, she provided essential information for identifying relevant decision criteria, defining the scoring scale, and performing the evaluation of residents. A total of 10 resident households . were selected for assessment, based on administrative recommendations and the availability of supporting data. Each household was evaluated across 14 criteria representing key poverty indicators, including housing conditions, access to clean water, income sources, education level, and asset ownership. The scoring for each criterion was assigned by Mrs. Nina based on field records and expert judgment, using a four-point ordinal scale in which lower scores indicate higher levels of deprivation and eligibility for social assistance. The complete evaluation results were compiled into a decision matrix that serves as the input for further processing using the AHP and SMART methods. Literature Review: The researcher conducted a literature review to deepen the understanding of the SMART (Simple Multi-Attribute Rating Techniqu. and AHP (Analytical Hierarchy Proces. methods used in this study. Additionally, the literature review aimed to explore concepts related to poverty, decision support systems, and relevant prior research. Design and Implementation of SMART and AHP Methods: Based on the results of problem analysis, data collection, and literature review, the researcher designed a decision support system by integrating the SMART and AHP methods. The AHP method was employed to determine the weight of each criterion, while the SMART method was used to calculate the final score for each aid recipient alternative based on the criterion scores. Report Writing: The final stage of this research involved compiling the report. The researcher documented all processes and results systematically, including the background, problem formulation, objectives, theoretical foundation, methodology, results and discussion, as well as conclusions and recommendations. This report is expected to serve as a reference for relevant stakeholders in making decisions regarding the distribution of social assistance. Figure 1. Research Framework Copyright A 2025 Authors. Page 43 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index 2 Research Data The criteria used in this decision support system were established based on field data collected through observation and The observation method involved direct examination of activities, conditions, and socio-economic realities within the Manyaran Subdistrict, while the interview method was conducted with Mrs. Nina Rizqiana Nugrahaeni. SE. Head of the Economic and Social Welfare Division at the Manyaran Subdistrict Office. Mrs. Nina Rizqiana Nugrahaeni. SE, also served as the designated domain expert responsible for providing judgments in the pairwise comparison process of the Analytic Hierarchy Process (AHP). Her expertise and institutional knowledge formed the basis for determining the relative importance of criteria and assessing the eligibility of residents for poverty-related social assistance. This study applies fourteen decision criteria, derived from national poverty indicators and adapted to local conditions in Manyaran. All criteria are considered benefit-type, meaning that higher values indicate more favorable or desirable attributes when evaluating aid eligibility. Each candidate . was assessed on a four-point ordinal scale, where a lower score reflects a higher level of deprivation and thus higher eligibility for aid. The full list of criteria, including their type, brief definitions, and scoring interpretations, is shown in Table 1. Table 1. Criteria Type, and Scoring Interpretation Code Criteria Type Score 4 (Very Does Not Mee. Score 3 (Does Not Mee. C01 Floor area of the residential building per person Type of residential Type of residential Sanitary facility Benefit < 4 mA per person 4Ae6 mA per person Benefit Earth Benefit Bamboo/rumbia Cheap bamboo/wood Low-quality wood Benefit No facility C05 Source of household lighting Benefit No electricity C06 Source of drinking Benefit River/unprotected Unprotected well/spring water C07 Cooking fuel Benefit Firewood Charcoal/kerosene C08 Frequency of meat/milk/chicken Purchase of new clothes per year Frequency of meals per day Benefit Less than once a Once a week Benefit Never 1 set of clothes Benefit 1 time 2 times Ability to pay for medical expenses Source of income of the household head Benefit Not able at all Benefit Highest education of the household Ownership of assets/savings Benefit Farm laborer/fisherman/land farmer (<500 mA with income \ 8 mA per person Ceramic/marble Plastered/painted Available and Stable electricity from PLN Piped water/bottled drinking water Non-subsidized gas/electricity Every day More than 3 sets of clothes 3 times with sufficient side Can pay without Employee/regular salary above UMR Completed junior high school or Owns sufficient assets/savings A total of 10 resident households . were selected for assessment, based on administrative recommendations and the availability of supporting data. The candidate . data is shown in Table 2. Copyright A 2025 Authors. Page 44 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index Table 2. Candidate (Alternativ. Data Code P01 P02 P03 P04 P05 P06 P07 P08 P09 P10 Full Address JL. GEDONG SONGO TIMUR 4 RT. 001 RW. JL. GEDONGSONGO TIMUR RT. 001 RW. JL. WR SUPRATMAN KAV 42 RT. 001 RW. JL GEDONGSONGO TMR IV/9 RT. 001 RW. GEDONGSONGO TIMUR IV RT. 001 RW. GEDONGSONGO TIMUR RT. RW. TMN. GEDONG SONGO TIMUR RT. 002 RW. JL. TMN GEDONGSONGO TIMUR 10 RT. 002 RW. JL GEDONGSONGO TIMUR RT. 003 RW. JL. TMN. GEDONGSONGO TIMUR RT. 003 RW. Name SUDARSONO Date Of Birth 1952-06-15 Place Of Birth SEMARANG Occupation Private Employee SRIYATUN 1959-06-02 SUKOHARJO Private Employee DJUMIATI 1949-02-18 BANDUNG Private Employee MARTINI HARSI 1960-07-29 KLATEN Private Employee SURIPAH 1955-12-31 SEMARANG Private Employee PUPON 1944-12-31 DEMAK HARTINI 1960-09-02 SEMARANG Farm/Plantation Laborer Private Employee HARNANIK 1957-04-24 SEMARANG Private Employee JUMENO 1958-01-02 SEMARANG SUKIRIN HS 1952-10-10 SRAGEN Farm/Plantation Laborer Farm/Plantation Laborer The results of the alternative evaluations for each criterion are shown in the following Table 3: Table 3. Alternative Evaluation Data Resident Code P01 P02 P03 P04 P05 P06 P07 P08 P09 P10 C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 3 Analytic Hierarchy Process (AHP) The Analytic Hierarchy Process (AHP) method is one of several methods used to solve Multi Attribute Decision Making (MADM) problems . Meanwhile. Multi Attribute Decision Making is the process of evaluating alternatives against a set of attributes or criteria, where each attribute is independent of the others . The AHP procedure in this research consists of the following steps. Creating a Hierarchy: A complex system can be understood by breaking it down into several supporting elements, arranging them in a hierarchy, and integrating them. Assessing Criteria and Alternatives: Criteria and alternatives are assessed through pairwise comparisons. For various problems, the scale from 1 to 9 is considered the best for expressing judgments. The levels of importance are shown in the following table . Table 4. AHP Scale of Importance Intensity 2,4,6,8 Description Both elements are equally important One element is slightly more important than the other One element is more important than the other One element is clearly more important than the other One element is absolutely more important than the other Intermediate values between adjacent judgments Summing the Values in Each Column of the Matrix Dividing Each Value in the Column by the Total Column Value to Obtain a Normalized Matrix. The formula for normalizing each value in a column is shown as follows . Copyright A 2025 Authors. Page 45 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index ycaycnyc Ocycn=1 ycaycnyc Evidence: a = Pairwise comparison matrix i = Row index of matrix a j = Column index of matrix a Summing the Values of Each Matrix Row and Dividing by the Number of Elements to Obtain the Average Value, using the formula. ycycn = ycu Ocycuyc=1 ycaycnyc Evidence: n = Number of criteria i = Average of the i-th row To ensure that the pairwise judgments are consistent, the Consistency Index (CI) and Consistency Ratio (CR) are computed . yaya = UUycoycaycuOeycu ycuOe1 yaycI = ycIycn Evidence: UUycoycaycu = principal eigenvalue of the comparison matrix = number of criteria = Random Index for a given ycu, based on standard AHP tables. A Consistency Ratio (CR) less than 0. 10 is considered acceptable, indicating that the pairwise comparisons are sufficiently consistent to be reliable. Random Index based on standard AHP tables is shown in Table 5. Table 5. Random Index Based on Standard AHP Ordo Matriks . Random Index (RI) 4 SMART (Simple Multi Attribute Rating Techniqu. SMART (Simple Multi Attribute Rating Techniqu. is a multi-criteria decision-making method developed by Edward in 1977. This multi-criteria decision-making technique is based on the theory that each alternative consists of a number of criteria with certain values, and each criterion has a weight that indicates its relative importance compared to the others. The Simple Multi-Attribute Rating Technique (SMART) is used in this study as a ranking tool that evaluates each alternative . based on their scores for each criterion and the weights derived from AHP. The steps of the SMART method in this research applied are as follows: Scoring Alternatives Each of the 10 alternatives was scored on 14 criteria using a four-point ordinal scale. The scores were assigned by an expert based on documented household data. All criteria were treated as benefit criteria, meaning a lower score reflects higher eligibility for aid. Determining Utility Values Determine utility values by converting the raw criterion values for each alternative into standardized criterion values. These utility values depend on the nature of the criteria themselves. Cost Criteria. Criteria in which "a smaller value is more desirable. " These types of criteria are usually in the form of expenses that must be incurred. The formula for calculating utility values for cost criteria is as follows. Copyright A 2025 Authors. Page 46 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index ya Oeya cayc ) = yaycoycaycuOeyaycuycyc ycoycnycu Evidence: cayc ) = Utility value of the i-th criterion for alternative ycayc yaycoycaycu = Maximum value of the criterion yaycoycnycu = Minimum value of the criterion yaycuycyc = Actual value of the i-th criterion . Benefit Criteria. Criteria in which "a larger value is more desirable. " These types of criteria are typically associated with benefits or gains. The formula for calculating utility values for benefit criteria is as follows . ya Oeya cayc ) = ya ycuycyc Oeyaycoycnycu ycoycnycu Evidence: cayc ) = Utility value of the i-th criterion for alternative ycayc yaycoycaycu = Maximum value of the criterion yaycoycnycu = Minimum value of the criterion yaycuycyc = Actual value of the i-th criterion . Determining the Final Score The final score of each alternative is calculated as the weighted sum of utility values, using the AHP-derived weights. caycn ) = Ocyco ya=ycn ycycn ycycn . caycn ) . Evidence: yc ( ycaycn ) : Total score of the alternative ycycn : Normalized weight of the i-th criterion ycycn . caycn ) : Utility value of the i-th criterion for the alternative The resulting scores are then used to rank the alternatives, and the top five households with the highest total scores are recommended as aid recipients. RESULT AND DISCUSSION In developing the Decision Support System for Social Assistance Provision for Poverty in Manyaran Subdistrict. Semarang, the researcher used the SMART and AHP methods as the model. The AHP method was used to calculate the importance weights of the criteria and to minimize the subjectivity in weighting due to the administratorAos evaluations. Once the inter-criteria weighting values were obtained from the AHP calculations, the next step was to rank the alternative data using the SMART method based on the values provided by the user for each alternative data. 1 Calculation of Criteria Weights Using the AHP Method The determination of the criteria priority weights was conducted by assigning values to each criterion. The evaluation data between criteria were obtained from interview with Mrs. Nina Rizqiana Nugrahaeni. SE, who serves as the Head of the Economic and Social Welfare Division at the Manyaran Subdistrict Office. As a subject matter expert, she provided essential information for identifying relevant decision criteria, defining the scoring scale, and performing the evaluation of residents. The evaluation rules between criteria using the Analytical Hierarchical Process (AHP) method follow those outlined in Table 1. Next, the criteria evaluation results were transformed into an evaluation matrix. The process of converting the criteria evaluation data into an evaluation matrix is performed by comparing the values of all criteria, including comparisons with the criteria itself. The comparison of identical criteria must have a value of 1, while the comparison of the criterion in the column to that in the row is the reciprocal ycA where ycA is the value for the criterion in the row corresponding to the column. The criteria evaluation matrix is shown as follows: Table 6. Criteria Evaluation Matrix Code C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 Copyright A 2025 Authors. Page 47 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index Code C12 C13 C14 Row Total C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 The next step is the process of normalizing the matrix, which is done by dividing each element of the criteria evaluation matrix by the row total. The row total is obtained by summing all values of each criteria column. Matrix normalization is performed using the formula of Equation . The normalized criteria results can be seen as follows: Table 7. Normalized Criteria Table Code C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 The next step is the calculation of the Priority Weights by summing the values from each row of the normalized matrix and dividing by the number of elements to obtain the average value. The calculation of the Criteria Priority Weights is performed using the formula in Equation . The result of the Criteria Priority Weights calculation is shown as follows: Table 8. Criteria Priority Weights Calculation Criteria Code C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 Criteria Name Floor area of the residential building per person Type of residential floor Type of residential wall Sanitary facility Source of household lighting Source of drinking water Cooking fuel Frequency of consuming meat/milk/chicken Purchase of new clothes per year Frequency of meals per day Ability to pay for medical expenses Source of income of the household head Highest education of the household head Ownership of assets/savings Priority Weight Type Benefit Benefit Benefit Benefit Benefit Benefit Benefit Benefit Benefit Benefit Benefit Benefit Benefit Benefit The final step is to ensure that the pairwise judgments are consistent. Based on the analysis of the pairwise comparison matrix consisting of 14 criteria, the maximum eigenvalue was found to be 14. This value was then used to calculate the Consistency Index (CI) using Equation . , resulting in a CI value of 0. Subsequently, referring to the standard Random Index (RI) value of 1. 57 for 14 criteria . s shown in Table . , the Consistency Ratio (CR) was computed using Equation . , yielding a CR value of 0. Since the resulting CR is below the acceptable threshold of 10, it can be concluded that the comparison matrix demonstrates an acceptable level of consistency. Therefore, the judgments in this matrix can be considered valid and reliable for use in the decision-making process employing the Analytical Hierarchy Process (AHP) method. 2 Ranking Alternatives Using the SMART Method The next step is to evaluate the alternatives against the criteria. In this ranking calculation, the SMART algorithm is applied using data from Social Assistance recipients for poverty in Manyaran Subdistrict. Semarang, consisting of a total of 10 households. The calculation is performed based on the criteria and the weight values for each criterion. Based on Table 3, it is observed that the minimum value is 1 and the maximum value is 4 for each criterion. The next step is to calculate the Utility value. Utility values are needed during the ranking of each alternative so that it can be determined which alternative is eligible to be selected. The Utility value for each alternative is calculated using Equation . for cost criteria and Equation . for benefit criteria. The utility calculation for each alternative is shown below: Copyright A 2025 Authors. Page 48 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index Utility Values for Alternative 1 2Oe1 1Oe1 4Oe1 2Oe1 2Oe1 ya01 = 4Oe1 = 0. 3333 ya02 = 4Oe1 = 0 ya03 = 4Oe1 = 1 ya04 = 4Oe1 = 0. 3333 ya05 = 4Oe1 = 0. ya06 = 1Oe1 4Oe1 = 0 ya07 = 4Oe1 4Oe1 = 1 ya08 = 2Oe1 4Oe1 4Oe1 = 1 ya09 = 2Oe1 2Oe1 4Oe1 = 0. 3333, ya10 = 2Oe1 3Oe1 4Oe1 = 0. 2Oe1 ya11 = 4Oe1 = 0. 3333 ya12 = 4Oe1 = 0. 3333 ya13 = 4Oe1 = 0. 3333 ya14 = 4Oe1 = 0. Utility Values for Alternative 2 1Oe1 4Oe1 1Oe1 1Oe1 1Oe1 2Oe1 ya01 = 4Oe1 = 0 ya02 = 4Oe1 = 1 ya03 = 4Oe1 = 0 ya04 = 4Oe1 = 0 ya05 = 4Oe1 = 0 ya06 = 4Oe1 = 0. 3Oe1 4Oe1 3Oe1 3Oe1 3Oe1 1Oe1 1Oe1 ya07 = 4Oe1 = 0. 6667 ya08 = 4Oe1 = 1 ya09 = 4Oe1 = 0. 6667 ya10 = 4Oe1 = 0 ya11 = 4Oe1 = 0 3Oe1 ya12 = 4Oe1 = 0. 6667 ya13 = 4Oe1 = 0. 6667 ya14 = 4Oe1 = 0. Utility Values for Alternative 3 1Oe1 1Oe1 4Oe1 4Oe1 1Oe1 3Oe1 ya01 = 4Oe1 = 0 ya02 = 4Oe1 = 0 ya03 = 4Oe1 = 1 ya04 = 4Oe1 = 1 ya05 = 4Oe1 = 0 ya06 = 4Oe1 = 0. 3Oe1 4Oe1 2Oe1 2Oe1 1Oe1 4Oe1 4Oe1 ya07 = 4Oe1 = 0. 6667 ya08 = 4Oe1 = 1 ya09 = 4Oe1 = 0 ya10 = 4Oe1 = 1 ya11 = 4Oe1 = 1 2Oe1 ya12 = 4Oe1 = 0. 3333 ya13 = 4Oe1 = 0. 3333 ya14 = 4Oe1 = 0. Utility Values for Alternative 4 3Oe1 1Oe1 1Oe1 1Oe1 2Oe1 3Oe1 3Oe1 3Oe1 2Oe1 1Oe1 1Oe1 ya01 = 4Oe1 = 0. 6667 ya02 = 4Oe1 = 0 ya03 = 4Oe1 = 0 ya04 = 4Oe1 = 0 ya05 = 4Oe1 = 0 2Oe1 ya06 = 4Oe1 = 0. 3333 ya07 = 4Oe1 = 0. 6667 ya08 = 4Oe1 = 0. 6667 ya09 = 4Oe1 = 0. 2Oe1 1Oe1 ya10 = 4Oe1 = 0. 6667 ya11 = 4Oe1 = 0. 3333 ya12 = 4Oe1 = 0 ya13 = 4Oe1 = 0,3333 ya14 = 4Oe1 = 0 Utility Values for Alternative 5 1Oe1 4Oe1 2Oe1 4Oe1 2Oe1 ya01 = 4Oe1 = 0 ya02 = 4Oe1 = 1 ya03 = 4Oe1 = 0. 3333 ya04 = 4Oe1 = 1 ya05 = 4Oe1 = 0. 2Oe1 4Oe1 4Oe1 2Oe1 3Oe1 ya06 = 4Oe1 = 0. 3333 ya07 = 4Oe1 = 1 ya08 = 4Oe1 = 1 ya09 = 4Oe1 = 0. 3333 ya10 = 4Oe1 = 0. ya11 = 1Oe1 4Oe1 = 0 ya12 = 4Oe1 4Oe1 = 1 ya13 = 1Oe1 4Oe1 = 0 ya14 = 3Oe1 4Oe1 = 0. Utility Values for Alternative 6 3Oe1 3Oe1 3Oe1 1Oe1 4Oe1 3Oe1 4Oe1 ya01 = 4Oe1 = 0. 6667 ya02 = 4Oe1 = 0. 6667 ya03 = 4Oe1 = 1 ya04 = 4Oe1 = 0. 6667 ya05 = 4Oe1 = 1 1Oe1 2Oe1 1Oe1 ya06 = 4Oe1 = 0. 6667 ya07 = 4Oe1 = 0 ya08 = 4Oe1 = 0 ya09 = 4Oe1 = 0. 3333 ya10 = 4Oe1 = 0 ya11 = 2Oe1 4Oe1 = 0. 3333 ya12 = 1Oe1 4Oe1 = 0 ya13 = 4Oe1 4Oe1 = 1 ya14 = 1Oe1 4Oe1 Utility Values for Alternative 7 2Oe1 2Oe1 3Oe1 1Oe1 3Oe1 3Oe1 3Oe1 4Oe1 1Oe1 ya01 = 4Oe1 = 0. 3333 ya02 = 4Oe1 = 0. 3333 ya03 = 4Oe1 = 0. 6667 ya04 = 4Oe1 = 1 ya05 = 4Oe1 = 0 1Oe1 3Oe1 3Oe1 ya06 = 4Oe1 = 0. 6667 ya07 = 4Oe1 = 0 ya08 = 4Oe1 = 0 ya09 = 4Oe1 = 0. 6667 ya10 = 4Oe1 = 0. 3Oe1 1Oe1 ya11 = 4Oe1 = 0. 6667 ya12 = 4Oe1 = 0. 6667 ya13 = 4Oe1 = 0. 6667 ya14 = 4Oe1 = 0 Utility Values for Alternative 8 3Oe1 4Oe1 2Oe1 3Oe1 3Oe1 3Oe1 3Oe1 1Oe1 ya01 = 4Oe1 = 0. 6667 ya02 = 4Oe1 = 1 ya03 = 4Oe1 = 0. 3333 ya04 = 4Oe1 = 0. 6667 ya05 = 4Oe1 = 0. 1Oe1 3Oe1 ya06 = 4Oe1 = 0 ya07 = 4Oe1 = 0. 6667 ya08 = 4Oe1 = 0. 6667 ya09 = 4Oe1 = 0. 6667 ya10 = 4Oe1 = 0 3Oe1 2Oe1 1Oe1 2Oe1 ya11 = 4Oe1 = 0. 6667 ya12 = 4Oe1 = 0. 3333 ya13 = 4Oe1 = 0 ya14 = 4Oe1 = 0. Utility Values for Alternative 9 Copyright A 2025 Authors. Page 49 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index 1Oe1 3Oe1 2Oe1 1Oe1 4Oe1 ya01 = 4Oe1 = 0 ya02 = 4Oe1 = 0. 6667 ya03 = 4Oe1 = 0. 3333 ya04 = 4Oe1 = 0 ya05 = 4Oe1 = 1 3Oe1 1Oe1 3Oe1 2Oe1 4Oe1 1Oe1 4Oe1 ya06 = 4Oe1 = 0. 6667 ya07 = 4Oe1 = 0 ya08 = 4Oe1 = 1 ya09 = 4Oe1 = 0 ya10 = 4Oe1 = 1 4Oe1 1Oe1 ya11 = 4Oe1 = 0. 6667 ya12 = 4Oe1 = 0. 3333 ya13 = 4Oe1 = 1 ya14 = 4Oe1 = 0 Utility Values for Alternative 10 3Oe1 4Oe1 1Oe1 1Oe1 1Oe1 ya01 = 4Oe1 = 0. 6667 ya02 = 4Oe1 = 1 ya03 = 4Oe1 = 0 ya04 = 4Oe1 = 0 ya05 = 4Oe1 = 0 4Oe1 4Oe1 4Oe1 4Oe1 4Oe1 2Oe1 2Oe1 4Oe1 ya06 = 4Oe1 = 0 ya07 = 4Oe1 = 1 ya08 = 4Oe1 = 1 ya09 = 4Oe1 = 0. 3333 ya10 = 4Oe1 = 1 4Oe1 ya11 = 4Oe1 = 1 ya12 = 4Oe1 = 1 ya13 = 4Oe1 = 0. 3333 ya14 = 4Oe1 = 1 The next step is to calculate the final score by multiplying the priority weight obtained from the AHP calculation results with the utility value for each attribute. The final score for each alternative is calculated using Equation . The final score for each alternative is shown in the following table: Table 9. Final Score Calculation for Each Alternative Code C01 C02 C03 C04 C05 C06 C07 C08 C09 C10 C11 C12 C13 C14 P01 P02 P03 P04 P05 P06 P07 P08 P09 P10 Total Score Based on the SMART-AHP calculation, five households with the highest eligibility for social assistance are identified as follows: Alternative P06 achieved the highest score of 0. 574, indicating the most favorable combination of deprivation indicators across all criteria. This is followed by P08 with a score of 0. 558, and P05 with 0. 524, both of which also exhibit strong eligibility characteristics. P10 ranks fourth with a score of 0. 515, while P03 completes the top five These results suggest that these five households demonstrate the highest levels of need and are therefore the most appropriate recipients for aid allocation according to the applied decision model. The residents or alternatives that place within the top 5 rankings are declared eligible to receive Social Assistance for Poverty in Manyaran Subdistrict. Semarang, whereas those outside the top 5 are declared ineligible. To assess the validity of the decision support system (DSS), a comparative analysis was conducted between the system-generated ranking and expert judgment provided by Mrs. Nina Rizqiana Nugrahaeni. SEAiHead of the Economic and Social Welfare Division at Manyaran Subdistrict. The top five alternatives identified by the system (P06. P08. P05. P10, and P. were cross-checked with the expertAos expectations based on her field knowledge and records. Table 10. Comparison of System vs Expert Rankings Alternative P02 P03 P05 P06 P08 P10 System Top 5 Yes Yes Yes Yes Yes Expert Top 5 Yes Yes Yes Yes Yes Match Yes Yes Yes Yes The results show a high degree of alignment with 80% Accuration. According to Mrs. Nina, households P06. P08, and P05 were indeed considered among the most critical cases requiring immediate social support. While there were slight differences in the ordering, four out of five households selected by the system were consistent with the expertAos top Therefore, the system demonstrated a strong level of accuracy and reliability in reflecting expert opinion, suggesting that the applied AHP-SMART model is both valid and appropriate for the selection of aid recipients. Among all alternatives. Alternative P06 achieved the highest final score . This household consistently showed high levels of deprivation across key criteria such as income source, education level, sanitation facilities, and asset ownership, which significantly influenced the utility scores when combined with the AHP-derived weights. For instance, the household's lack of stable income and inadequate living conditions resulted in higher normalized utilities for highly weighted criteria like C01 . loor are. C03 . all typ. , and C04 . Copyright A 2025 Authors. Page 50 This Journal is licensed under a Creative Commons Attribution 4. 0 International License Bulletin of Informatics and Data Science Vol. 4 No. May 2025. Page 42Oe52 ISSN 2580-8389 (Media Onlin. DOI 10. 61944/bids. https://ejurnal. id/index. php/bids/index In contrast, the lowest-ranked household (P. obtained a total score of only 0. This alternative generally exhibited lower levels of deprivation, with relatively better housing conditions, basic utilities, and income indicators. The large score gap between P06 and P04 highlights the modelAos ability to distinguish levels of eligibility based on multidimensional poverty indicators. A comparative pattern emerges when analyzing the top five alternatives (P06. P08. P05. P10. versus the remaining five: Top-ranked households tend to lack access to proper housing, stable income, healthcare affordability, and educationAi often scoring 3 or 4 on most criteria. In contrast, lower-ranked households typically have more stable conditions in one or more critical areas . , better construction, income source, or access to utilitie. , leading to lower utility values and thus lower final scores. These findings support the robustness of the model in capturing the multidimensional nature of poverty and prioritizing those most in need. Although the DSS model performs well, several limitations must be acknowledged: Equal Pairwise Judgments: While the AHP structure was applied, the initial pairwise comparisons used relatively consistent and symmetric values. Involving multiple experts or stakeholders could improve the granularity and credibility of the weight assignment process. Static Weighting Assumptions: The weights derived from AHP are context-specific and static. Future models can integrate dynamic weighting or fuzzy AHP to better reflect uncertainty or variability in expert judgments. Limited Data Scope: The current study only evaluated 10 households due to practical constraints. Applying the model to a larger dataset would improve its generalizability and test its scalability. No Consideration of Temporal Factors: Changes in household conditions over time are not considered. A longitudinal approach or integration with real-time data sources could enhance decision responsiveness. Future development may include building an interactive DSS interface for local governments, integrating realtime data input from field officers, or combining machine learning techniques with MCDM for pattern recognition and classification of aid eligibility. CONCLUSION A Decision Support System for the distribution of poverty social assistance has been developed in the Manyaran Subdistrict. Semarang, using the SMART (Simple Multi Attribute Rating Techniqu. and AHP (Analytical Hierarchy Proces. A total of 14 poverty-related criteria were established, and 10 candidate households were evaluated through expert scoring. The AHP method was used to determine the relative weight of each criterion, while the SMART method provided a final ranking of alternatives based on their normalized scores and the derived weights. The system successfully identified the five most eligible households as P06. P08. P05. P10, and P03, with P06 obtaining the highest score due to significant levels of deprivation in several high-priority criteria such as income source, housing conditions, and education level. In contrast, the lowest-ranked household. P04, showed comparatively better conditions in multiple key areas. Overall, the top five households tended to share characteristics such as poor structural housing, limited or unstable income, low education attainment, and lack of access to sanitation or savingsAifeatures less evident in the lowerranked alternatives. To validate the reliability of the system, the results were compared with expert judgment from Mrs. Nina Rizqiana Nugrahaeni. SE, the Head of the Economic and Social Welfare Division at the Manyaran Subdistrict Office. The DSS produced a top-five ranking that matched four out of five of the expertAos top priorities, achieving an alignment accuracy of 80%. This indicates that the systemAos recommendations are largely in agreement with domain expertise, reinforcing its validity and practical relevance. Despite these promising outcomes, the study is not without The evaluation was restricted to a small sample of 10 households, which limits the generalizability of the Furthermore, while AHP was used to determine criterion weights, the pairwise comparisons applied relatively uniform values, limiting the method's ability to reflect nuanced expert preferences. Additionally, the study relied on a single evaluator for both scoring and weight determination. Future work should address these limitations by involving multiple experts to refine the weighting process, expanding the number of evaluated households, and conducting realworld trials to measure the systemAos effectiveness in practice. It is also recommended to develop a user-friendly DSS interface for use by local decision-makers, explore comparisons with other MCDM techniques, and incorporate dynamic or real-time data inputs to better reflect changing socio-economic conditions. In conclusion, the integration of AHP and SMART in this DSS framework offers a structured, transparent, and replicable approach to aid recipient selection. With further development, the system has strong potential to support more objective, equitable, and evidence-based social assistance programs at the community level. REFERENCES