Protective effects of dendropanax morbifera against cisplatin-induced nephrotoxicity without blocking chemotherapeutic efficacy in animal models

Kim, Jisu Sungkyunkwan university 2019 국내석사

Cisplatin (CDDP)은 광범위한 고형 종양들의 치료에 널리 사용되는 화학 요법제 이다. 하지만, 많은 환자에서 발생하는 급성 신장 손상 (AKI)에 의해 임상적 사용이 제한된다. 집중적인 연구에도 불구하고, Cisplatin 유도성 AKI에 대한 성공적으로 보호하는 치료법은 없다. 본 연구의 목적은 Dendropanax morbifera (DM)의 Cisplatin 유도 성 AKI에 대한 신장 보호 효과를 조사하고 cisplatin 화학 요법 동안 효과적으로 표적화 될 수 있는 효과를 조사하는 것이었다. 실험 디자인에서 5일간 실험을 하였으며, 4 마리의 수컷 Sprague-Dawley 쥐 그룹; Control; Cisplatin (6 mg / kg, i.p.); DM (25 mg / kg, p.o.) 과 복합군으로 DM (25 mg / kg, p.o.)을 Cisplatin 주사 2 시간 전에 사용 하였다. 본 연구에서는 cisplatin의 투여로 신장의 정상 조직 학적 구조 변화와 함께 체중 감소, 혈액 요소 질소 (BUN) 및 크레아티닌 증가와 TNF- α, IL-1β, IL-6, IL-10 를 포함한 염증 유발 사이토 카인 수준이 감소되었다. SBP1, KIM-1, NGAL 및 TIMP-1과 같은 단백질 기반 신 독성 생체 표지 물질의 요로 배설 또한 Cisplatin 군에서 증가하였다. 반대로, DM은 신장 손상 생체 표지 물질 (BUN, 크레아티닌 및 KIM-1, NGAL 및 SBP1)의 현저한 감소로 명백한 Cisplatin 유도 신장 독성을 유의하게 보호했다. DM 치료는 내인성 항산화제 활성 (SOD 및 Catalase)을 증가시킴으로써 신장에서 Cisplatin 유도된 산화 스트레스를 현저하게 감소시켰다. 또한, DM 치료는 또한 염증성 사이토 카인의 수준을 감소시켰다. 특히, DM의 보호 효과는 조직 병리학적 검사에서 명확하게 관찰되었으며, DM 처리로 인한 신장은 Cisplatin에 의해 유발된 근위 세뇨관에서의 심한 신장 손상을 현저하게 감소시켰다. 종양 이종 이식 모델에서 DM은 형질 전환된 대장암 세포에 Cisplatin 매개 항암 활성에 영향을 미치지 않았으나, Cisplatin의 화학 요법 활성을 향상시켰을 뿐만 아니라 Cisplatin 유도성 AKI에 대한 보호 효과를 나타냈다. 종합적으로 말하자면, 이러한 결과는 Cisplatin에 의한 신 독성에서 DM의 보호적인 역할을 보여주며 Cisplatin에 기초한 암 치료 중에 renoprotective agent에 사용되는 신뢰할 수 있는 전략이다. Cisplatin (CDDP) is generally used chemotherapeutic material for treatment of various solid tumors. However, its clinical use is limited by occurs acute kidney injury (AKI) in many patients. Despite intensive research, there is no successful protective therapy against CDDP-induced AKI. The purpose of the present study was to demonstrate renoprotective effects of Dendropanax Morbifera (DM) on CDDP-induced AKI and which can be effectively targeted during CDDP chemotherapy. In the experimental design, four groups of male Sprague-Dawley rats; Control (vehicle); CDDP (6 mg/kg, i.p.); DM (25 mg/kg, oral) for 5 days; and DM (25 mg/kg, oral) 2 h before CDDP injection. In the present research, injection of CDDP decreased body weight, but increased blood urea nitrogen (BUN) and creatinine and pro-inflammatory cytokine levels including TNF- α, IL-1β, IL-6, IL-10 along with alteration in normal histological architecture of kidney. Urinary nephrotoxicity biomarkers like selenium-binding protein 1 (SBP1), kidney injury molecule-1 (KIM-1), neutrophil gelatinase-associated lipocalin (NGAL), and tissue inhibitor of metalloproteinase-1 (TIMP-1) also increased in the CDDP-treated group. On the contrary, DM significantly protected CDDP-induced nephrotoxicity which was evident by significant reduction of renal injury biomarkers (BUN, creatinine, and KIM-1, NGAL, and SBP1). DM treatment markedly reduced CDDP-induced oxidative stress in kidney by activating endogenous antioxidants activities (SOD and catalase). Further, DM treatment also reduced the levels of pro-inflammatory cytokines. In particular, protective effect of DM was clearly observed in histopathological examination wherein, kidneys from DM treatment markedly reduced CDDP-induced severe kidney damages in the proximal tubules. In tumor xenograft model, DM did not affect CDDP-mediated anticancer activity in transfected colon cancer cells, but enhanced the chemotherapeutic activity of CDDP as well as exhibited protective effects on CDDP-induced AKI. These results demonstrate protective effects of DM in CDDP-induced nephrotoxicity and suggest as a reliable strategy used for renoprotective agent during CDDP-based cancer therapy.

Nonresponse weighting adjustment using nonparametric response propensity score

김지수 Graduate School, Yonsei University 2005 국내석사

In the survey analysis, nonresponse leads to many problems. First, nonreponse reduces the effective sample size, so this increases the variance of the resulting estimator. Moreover, nonresponse incurs nonresponse bias which is the propensity of non-respondent is not the same as that of respondent.To prevent this nonresponse bias, many researchers use various nonresponse weighting adjustment(NWA) methods with estimated response propensity score(RPS). One of the most popular NWA methods is the logistic regression for estimating RPS.If the response structure follows the logistic distribution, estimated statistic(ex. Interested mean, total, variance etc.) satisfies unbiasedness. However, if not, estimated statistic may suffer great bias problem. So, some remedy is required for this case.This article will suggest the local smoother which is a kind of the nonparametric methods to resolve the bias problem. LOCFIT procedure will be introduced as the local smoothing method and extended to GAM(Generalized Additive Model).

Dracocephalum Palmatum Stephan Extract Induces Caspase- and Mitochondria-dependent Apoptosis via MYC Inhibition in Diffuse Large B-cell Lymphoma

김지수 부산대학교 대학원 2020 국내석사

Dracocephalum palmatum Stephan (DPS), a medicinal plant used by Russian nomads, has been known to have anti-oxidant properties. Here, we tested the tumor-suppressive effect of DPS extract (DPSE) in diffuse large B-cell lymphoma (DLBCL). Exposure of multiple DLBCL cell lines to DPSE significantly decreased cell viability and increased apoptosis, whereas it had no effect on the survival of normal cells in vitro and in vivo. This indicates that its cytotoxic effect may be specific to cancer cells. Mechanistically, cell death induced by DPSE was dependent on the activation of caspase-3/7 and the disruption of mitochondrial membrane potential. Treatment with the extract ameliorated the expression of anti-apoptotic Bcl-2 members Bcl-xL and Mcl-1, and upregulated that of pro-apoptotic Bcl-2 members Bax and Bak. These modulations led to the disruption of mitochondrial membrane potential that culminated in the activation of executioner caspases-3 and 7. Intriguingly, Myc overexpression blocked DPSE-induced cell killing, indicating the involvement of Myc in the process. Given that dysregulation of Myc is strongly associated with the pathobiology of DLBCL, our study highlights the potential therapeutic efficacy of DPSE in DLBCL patients with aberrant Myc expression. Further, fractionation of DPSE by thin layer chromatography (TLC) and liquid chromatography (LC)/mass spectrometry (MS) -based investigation of the fraction with bioactive compounds demonstrated that flavonoids may be responsible for most, if not all, anti-lymphoma effect. Efforts to identify the bioactive flavonoids is currently underway and it is expected to discover other new anti-cancer compounds.

Low Current Sensing Circuit for Spin Transfer Torque MRAM (STT-MRAM) : STT-MRAM을 위한 저전류 센싱회로

Jisu Kim 연세대학교 일반대학원 2015 국내박사

DRAM, SRAM, 플래시 메모리와 같은 기존 메모리들은 각각의 특성에 따라 다른 어플리케이션에 사용된다 [1], [2]. 이러한 기존 메모리들의 장점을 모두 가진 유니버셜 메모리를 구현하기 위해 차세대 비휘발성 메모리에 대한 연구가 활발히 진행되고 있다. STT-MRAM은 차세대 비휘발성 메모리의 주요한 후보이고 상대적으로 작은 비트셀 크기를 가지면서 SRAM 만큼 빠른 비휘발성 메모리이다 [3]. 또한 저전력으로 동작하며 CMOS 공정과 호환성이 높아 경제적이다. 그러나 직접도가 높은 미세공정에서 구현하기 위해 극복해야할 문제점이 존재한다. 그 중 하나는 STT-MRAM의 저장소자인 MTJ를 write 하기 위해 요구되는 쓰기 전류가 커서 미세공정에서 그 전류를 확보하기 어려운 것이다. 그러나 그것은 최근 perpendicular STT-MRAM의 개발로 쓰기 전류가 충분히 감소되어 해결되었다 [4]. 반면, 센싱 과정에서는 극복해야 할 문제들이 여전히 있다 [5]. 첫째, 공정이 미세화 됨에 따라 공정 변이가 증가하고 그것으로 인해 센싱 마진이 감소된다. 이것은 센싱을 어렵게 만들고 센싱 수율을 감소시킨다. 둘째, 공정이 미세화 됨에 따라 reliablilty와 동적 전력 소모를 감소시키기 위해 공급전압 (VDD)이 감소되는데 그것으로 인해 센싱 마진이 감소된다. 일반적으로 공정 미세화 과정에서 소자의 performance를 유기하기 위해서는 문턱 전압 (Vth)도 공급 전압과 같이 감소되어야 하는데 누설 전류가 증가하는 문제 때문에 문턱 전압은 공급 전압보다 덜 감소한다. 따라서, 공급전압과 문턱전압의 차이가 감소하고 그것으로 인해 센싱 마진이 감소하게 된다. 모바일 기기에서 이러한 현상은 두드러진다. 모바일 기기는 한정된 베터리로 구동되고 대부분의 동작을 대기모드로 동작하므로 누설전류를 감소시키기 위해 문턱전압이 큰 트랜지스터를 사용하는 것이 일반적이다. 따라서, 모바일 어플리케이션을 위한 STT-MRAM에서는 높은 문턱전압으로 인해 센싱 마진 확보가 더 어렵게 된다. 본 논문에서는 공정이 미세화 하는 과정에서 기존 센싱 회로의 센싱 수율이 감소되는 근본적인 원인을 분석하였고 그것을 기반으로 낮은 센싱 전류와 높은 센싱 수율을 갖는 source degeneration sensing circuit (SDSC)를 제안하였다. 65nm 공정의 모델 파라메터를 사용한 Monte carlo 시뮬레이션의 결과는 기존 센싱 회로는 공급전압 1.1V (1.6V) 에서 48.0 μA (64.2 μA) 의 센싱 전류, 0 % (81.5%) 의 센싱 수율을 갖지만 SDSC는 공급전압 1.1V에서 43.1 μA의 센싱 전류, 96.3 %의 센싱 수율을 갖는다는 것을 보여준다. 하지만, 45 nm 공정의 모델파라메터를 이용한 시뮬레이션에서는 공정 미세화로 인한 공급 전압의 감소와 공정 변이의 증가로 인해 SDSC의 센싱 수율이 0 %로 감소된다. 이 문제를 극복하기 위해 sensing circuit with self-body biasing (Self-BB) 와 split-path sensing circuit (SPSC) 가 제안되었다. Self-BB와 SPSC는 self-body 바이어스 기술과 가변 Vref 기술이 사용되어 SDSC 보다 큰 센싱 마진을 얻는다. 그 결과 Self-BB와 SPSC는 33 μA의 센싱 전류에서 각각 100.0 %와 96.3 %의 센싱 수율을 얻는다. 두 회로는 low-leakage 트랜지스터를 사용하고 간단한 구조를 사용해서 공정 미세화로 인해 야기되는 센싱 문제를 해결하였다. Conventional memories, such as DRAM, SRAM and flash memory, have unique merits and demerits and thus they are selected according to the application type. Next-generation non-volatile memories are being actively studied to realize a universal memory that uses the unique merits of each conventional memory [1-2]. The spin transfer torque MRAM (STT-MRAM) is a noticeable next-generation non-volatile memory that can operate as fast as SRAM with a relatively small bitcell size [3]. In addition, it consumes lower power and is compatible with the existing CMOS process. To improve the degree of its integration, however, many issues arising from technology scaling must be addressed. One of these issues is that it is difficult to achieve a sufficient write current caused by the large switching current (IC) for magnetic tunnel junction (MTJ). The recent advances in the perpendicular STT-MRAM technology, however, successfully proved the possibility of achieving a small IC, and paved the way for further technology scaling [4]. On the other hand, a few more critical issues in the sensing operation should be resolved for further technology scaling [5]. First, technology scaling increases the process variation, which in turn increases the sensing margin variation. This makes sensing difficult and thus reduces yield. Second, as the technology is scaled down, the supply voltage (VDD) decreases to guarantee reliability and to reduce the dynamic power consumption. To maintain the device performance, the threshold voltage (VTH) also decreases [6]. VTH, however, decreases at a lower rate than VDD to prevent increasing the leakage current, leading the degradation in the sensing margin. This tendency becomes more serious if handheld devices are considered. Handheld devices are driven by limited battery capacity, and are usually in the standby mode. Thus, a low-leakage transistor with a high VTH must be used to reduce the standby leakage current, and consequently, to maximize the battery lifetime [7]. The sensing margin of the STT-MRAM sensing circuit for mobile applications is then further reduced because of the higher VTH. In this dissertation, conventional sensing circuits have been analyzed to understand the root causes of the sensing yield degradation with process technology scaling. And, novel sensing circuit to obtain high yield with low sensing current, source degeneration sensing circuit (SDSC) is proposed. Monte Carlo HSPICE simulation results using 65nm technology model parameters show that SDSC achieves an read access yield of 96.3% with a sensing current of 43.1 μA at a supply voltage of 1.1V for 32M bit, whereas the conventional sensing circuit achieves an read access yield of only 0% (81.5%) with a sensing current of 48.0 μA (64.2 μA) at a supply voltage of 1.1V (1.6V). However, in the simulation using 45 nm technology model parameters, SDSC cannot obtain target read access yield because of reduced VDD and increased process variation by technology scaling. To overcome this problem, two sensing circuits, sensing circuit with self-body biasing (Self-BB) and split-path sensing circuit (SPSC) are proposed. Self-BB and SPSC have larger sensing margin than SDSC using self-body biasing and variable Vref scheme, respectively. Consequently, Self-BB and SPSC achieves an read access yield of 100.0% and 99.3% with the sensing current of 33 μA for 32M bit, respectively. They mitigate the sensing issues caused by technology scaling and the use of a low-leakage transistor with simple structure.

Role of critical micelle concentration on the oxidation stability in bulk oils using lecithin

Kim, Jisu Sungkyunkwan university 2018 국내석사

Critical micelle concentration (CMC) is defined as the concentration of amphiphilic compounds above which micelles are spontaneously formed. The CMC is important in food chemistry, because the association colloids formed in CMC play an important role in the oxidative stability of bulk oils. Therefore, the objectives of this study were to investigate effects of moisture content and amphiphilic compounds on the critical micelle concentration of lecithin (Hydrophile-Lipophile Balance, HLB 4.5) in bulk oils and to determine antioxidant or prooxidant properties of lecithin in corn oil by oxidation. CMC of lecithin was determined in bulk oils and in medium chain triacylglycerol (MCT) with different moisture content and degree of oxidation. CMC in MCT was significantly higher than other vegetable oils including olive, soybean, corn, and rapeseed oils (p<0.05). Presence of moisture significantly affected the CMC in MCT (p<0.05). CMC was high when moisture content was below 900 ppm whereas at 1000 ppm moisture content, CMC decreased significantly (p<0.05) and then started to increase. Addition of total polar materials (TPM), which are oxidation products, at 3 and 5% decreased CMC significantly (p<0.05) in MCT compared to MCT with 0, 1 and 1.5% TPM. As degree of oxidation increased in corn oil, CMC decreased gradually. Also, corn oils under different moisture content showed similar pattern of CMC in MCT whereas oxidized corn oil had a little lower CMC than unoxidized corn oil. The results clearly showed that concentration of lecithin for the formation of micelle is greatly influenced by the presence of oxidation products and moisture content in bulk oils. Effects of lecithin (HLB 4.5) on the oxidative stability were determined in corn oil at 50°C storage. Concentration of lecithin ranged from 100 to 4000 ppm, which covered the CMC in oils. Headspace oxygen content in corn oil containing lecithin decreased less with concentration dependent manner, which implies antioxidative properties of lecithin. Lecithin added samples near and over CMC had higher oxidative stability compared to others based on conjugated dienoic acid and p-anisidine value assays. Tocopherols in oils containing near CMC were detected in the highest concentration. Corn oil added with near CMC of lecithin showed the highest oxidative stability which could be due to the incorporation of tocopherols with association colloids formed by added lecithin. In conclusion, the factor affecting CMC is water content and total polar materials. Also, CMC is a very significant indicator of oxidative stability. Especially, oxidation stability was higher when the concentration of lecithin which is CMC point and over.

SLOWPIXELS : Slow-design for reflective retrieval of personal photos

Jisu Kim Graduate School of UNIST 2020 국내석사

We introduced the design and findings in our user study of SLOWPIXELS, a photo printer which aims to stimulate reflection & memory recall through the re-materialization of digital photos. Photos were serendipitously retrieved dependent upon position of two sliders: When did I take the photo (photo date slider) & Where did I take the photo (distance from home slider). SLOWPIXELS used its owner’s Instagram account metadata. We tested SLOWPIXELS to 20 participants for 11 days to investigate how the Slow-design could support reflective retrieval of personal photos. Findings revealed that SLOWPIXELS encourage users to reflect their past by unexpected pictures and promote their social bond. We also found that SLOWPIXELS has a design possibility in the public space for event-like situation. For example, in café or student community centre. These findings suggested several contributions, such as designing for revisiting serendipitous memories on the past, empowering social communication, and new way of retrieval of photography while exploring the slow-design agenda within the Design Research community.

The Design of Nano-Structured Materials for Superior Performance in Lithium-Ion and Aluminum-Ion batteries

Kim, Jisu 영남대학교 대학원 2022 국내석사

Fast-charging lithium-ion batteries (LIBs) with high energy and power density are important for the electric vehicle (EV) industry. However, graphite anode materials with high conductivity and stability still face setbacks in achieving fast charging and reach its potential. To improve the fast-charging ability of graphite, pristine graphite (PG) were treated with acid/base media, which break the bulk graphite layers into thin expanded graphite layers. Specifically, KOH etching is an effective method as it creates pores on the surfaces, facilitating the movement of Li+ ions and as well as increasing the number of sites for Li+ ion intercalation in the graphite layers. The specific capacities of acid-treated graphite (AG) and KOH-treated graphite (KG) is > 300 mAh g−1 at a current density of 100 mA g−1, which is higher than PG. Interestingly, even at high current densities, AG and KG exhibit good retention capacity in cycling. Furthermore, after long term stability tests, KG displayed the best recovery capacity at fast charging (1,000 mA g−1) and slow discharging (200 mA g−1) with the highest specific capacity of 250 Ah g−1. In summary, the fast-charging ability in LIBs can be improved using surface modified graphite as the anode material. Rechargeable aluminum-ion batteries (AIBs) are a new generation of low-cost and large-scale electrical energy storage systems. However, AIBs suffer from a lack of reliable cathode materials with insufficient intercalation sites, poor ion-conducting channels, and poor diffusion dynamics of large chloroaluminate anions (AlCl4− and Al2Cl7−). To address these issues, surface-modified graphitic carbon materials (i.e., acid-treated expanded graphite (AEG) and base-treated graphite (BEG)) are developed as novel cathode materials for ultra-fast chargeable AIBs. The AIB system employing AEG exhibits a specific capacity of 88.6 mAh g−1 (4 A g−1) and ≈80 mAh g–1 at an ultra-high current rate of 10 A g–1 (≈99.1% over 10000 cycles). Impressively, the AIB system based on the BEG cathode delivers a high specific capacity of 110 mAh g−1 (4 A g−1) and ≈91 mAh g–1 (≈99.9% over 10000 cycles at 10 A g–1). Moreover, the BEG cell has a high energy and power densities of 247 Wh kg–1 and 44.5 kW kg–1. This performance is one of the best among the AIB graphitic carbon materials reported to date for chloroaluminate anions storage and are significant for the further development of rechargeable AIBs with high energy, high power density, and exceptionally long life. 배터리는 화학적 에너지를 전기로 변환하여 전자를 생산하고 저장하는 장치이다. 리튬 이온 배터리는 충전하고 다시 사용가능한 이차전지로써, 리튬 이온이 전극과 전해질 사이를 이동하면서 충방전 하는 배터리이다. 이러한 리튬 이온 배터리의 성능을 향상시키기 위해 전극 소재의 연구와 개발이 활발히 진행되고 있다. 본 논문에서는 고속 충전에서 안정적인 셀 구동과 높은 용량을 나타내기 위해서 대표적인 음극 소재의 흑연을 표면 처리하였다. 손쉬운 표면처리로 흑연의 층간 간격과 기공을 만들어 리튬 이온의 이동 속도를 빠르게 해주었다. 그 결과로 다양한 전류밀도에서 높고 안정적인 성능이 관찰되었다. 특히, 1A/g의 빠른 전류에서 고용량을 나타냄을 관찰하였다. 다양한 산업에서 사용되는 리튬 이온 배터리는 화재의 위험성과 원자재 가격 상승으로 차세대 배터리의 개발이 필요하다. 그 중에 알루미늄 이온 배터리는 안전성이 높고, 풍부한 알루미늄 자원으로 가격이 저렴하다. 그리고 3전자 반응으로 높은 부피비 용량을 가지고 있다. 이러한 알루미늄 이온 배터리에서 염화 알루미늄 이온의 원활한 이동을 위해 흑연 양극재를 개선하였다. 개선된 흑연 양극재를 적용시킨 결과, 초고속 충전 속도인 10A/g에서 안정적인 셀 구동을 보였다. 또한, 개선된 흑연을 적용한 알루미늄 이온 배터리는 다른 양극재나 어플리케이션보다 높은 파워와 에너지 밀도를 가짐을 확인하였다. 간단한 방법으로 전극 소재를 처리하여 높은 용량과 안정적인 배터리의 성능을 이루어 내었다.

Heat generation in nano-pigments via photothermal effect of hollow au nanoparticles

Kim, Jisue Kwangwoon University Graduate School 2022 국내석사

The photothermal effect resulting from the localized surface plasmon resonance (LSPR) phenomenon of metal nanoparticles is essentially a light–heat conversion mechanism, wherein the heat is generated around the nanoparticles due to the absorption of light by the nanoparticles. A pigment with the photothermal effect can be used as a multifunctional nano-pigment providing high heat energy from a small amount of light energy as well as a bright color. In this study, a grayish blue nanopigment (GBNP) was synthesized using silica nanoparticles and hollow Au nanoparticles (HAuNP) that exhibited a relatively high photothermal effect in the near-infra red (NIR) region. By irradiating light in the NIR region, GBNP could selectively produce stable thermal energy with a photothermal efficiency of 39%. It was mixed with paint to confirm its applicability to pigments and heat-generation performance. The paint containing 4 wt% GBNP revealed blue-gray color, and the surface temperature of GBNP paint was reached to 110 °C, exhibiting a stable heatgeneration performance under both repeated and continuous NIR irradiation. In addition, the thermal energy generated by the photothermal effect implied the high evaporation efficiency of water droplets on the surface. This pigment might be utilized in applications where a selectively high temperature needs to be maintained continuously for bacterial-growth prevention or moisture control.

Co-precipitation synthesis of amorphous Ru-Ti oxides as bifunctional catalysts for Li-O2 batteries

Kim, Jisu Sungkyunkwan university 2017 국내석사

I know your stance! analyzing Twitter user political stance on diverse perspectives

Kim, Jisu Sungkyunkwan University 2022 국내석사

The popularity of social network service users has increased in recent years, altering politicians’ interest level in social network services. Given this trend, social network services now play a central role in political communication channels, enabling them to express and share their opinions and news directly with citizens. Therefore, many researchers have attempted to investigate the political stance of social network service users and have proposed a user's political stance model utilizing this dataset. Understanding and detecting social network services and a user's political stance can play a significant role in marketing strategies and determining election winners. In light of this, the present study examined Twitter from diverse perspectives to analyze and detect a Twitter user's political stance. This study collected Twitter datasets and labeled a user’s stance using a clustering approach to determine whether a user was a Democrat or a Republican. After an exploratory analysis of users' tweet content: image and text, user network, and profile description, the tweet user stance detection model was proposed and tested. The results indicated notable differences between Democrats and Republicans from diverse perspectives on Twitter, with an accuracy of 85.35% compared with baseline models. The implications and limitations of this study were discussed based on the results. 소셜 네트워크 서비스의 인기가 높아지면서 정치인들 또한 소셜 네트워크 서비스에 많은 관심을 갖게 되었다. 이러한 추세로, 소셜 네트워크 서비스는 정치적 소통 채널로서 중심적인 역할을 하고 있는데, 정치인들의 의견과 뉴스를 시민들과 직접 공유하고 소통할 수 있게 해주기 때문이다. 이로 인해 많은 연구자들이 소셜 네트워크 서비스 사용자의 정치적 입장을 조사하고 소셜 네트워크 서비스 데이터 세트로 사용자의 정치적 입장을 예측하는 모델을 제안하려고 시도한다. 소셜 네트워크 서비스를 이해하고 사용자의 정치적 성향을 파악하는 것은 선거 활동에 있어 마케팅 전략과 선거 승자를 결정하는데 중요한 역할을 할 수 있기 때문이다. 따라서 본 연구는 트위터 사용자의 정치적 입장을 분석하고 탐지하기 위해 다양한 관점에서 트위터를 조사한다. 따라서 이 연구는 트위터 데이터 세트를 수집하고 사용자가 민주적 성향인지 보수적 성향 인지 클러스터링 기법을 통해 레이블링을 진행하였다. 이후, 사용자의 트윗 컨텐츠: 이미지 및 텍스트, 사용자 네트워크 및 프로필 설명에 대한 탐색적 분석을 수행한 후 트윗 사용자 정치성향 탐지 모델을 제안하고 테스트한다. 결과는 트위터에 대한 다양한 관점에서 민주당 지지자와 공화당 지지자 사이의 현저한 차이가 있음을 나타내며, 모델은 기준 모델과 비교하여 85.35%의 정확도로 가장 잘 수행되었다. 마지막으로, 결과에 따라 시사점과 한계점에 대해 논의한다.