• Asia pacific journal of information systems
• /
• 제31권4호
• /
• pp.472-490
• /
• 2021

The detection of products with manipulated reviews has received widespread research attention, given that a truthful, informative, and useful review helps to significantly lower the search effort and cost for potential customers. This study proposes a method to recognize products with manipulated online customer reviews by examining the sequence of each review's sentiment, readability, and rating scores by product on randomness, considering the example of a Russian online retail site. Additionally, this study aims to examine the association between brand awareness and existing manipulation with products' reviews. Therefore, we investigated the difference between well-known and unknown brands' products online reviews with and without manipulated reviews based on the average star rating and the extremely positive sentiment scores. Consequently, machine learning techniques for predicting products are tested with manipulated reviews to determine a more useful one. It was found that about 20% of all product reviews are manipulated. Among the products with manipulated reviews, 44% are products of well-known brands, and 56% from unknown brands, with the highest prediction performance on deep neural network.

• 인터넷정보학회논문지
• /
• 제10권6호
• /
• pp.81-98
• /
• 2009

협업적 여과는 추천 시스템을 구축하는데 가장 널리 보급된 정보 여과 기법으로 사용자 각 개인의 관심에 적합한 정보 및 아이템을 추천함으로써 사용자들의 의사 결정에 도움을 준다. 그러나 협업적 여과 기법은 우수한 추천 성능에도 불구하고, 최근에는 실링 공격이라 일컫는 악의적인 목적을 가진 사용자들의 추천 결과 조작에 쉽게 노출될 수 있는 문제가 새로운 이슈로 대두되고 있다. 본 논문에서는 협업적 여과의 실링 공격 문제들을 보완하기 위해, 추천 시스템에서 발생할 수 있는 실링 공격의 유형을 분석하고 악의적인 사용자의 조작된 선호도가 시스템에 미치는 영향을 최소화하기 위한 강건한 신뢰 모델 구축 방법을 제시한다. 그리고 그 모델을 적용하여 신뢰할 수 있는 아이템 추천 및 선호도 예측 방법을 제안한다.

• 국제학술발표논문집
• /
• The 5th International Conference on Construction Engineering and Project Management
• /
• pp.437-443
• /
• 2013

Recently, sustainable building design, a growing field within architectural design, has been emerged in the construction industry as the practice of designing, constructing, and operating facilities in such a manner that their environmental impact, which has become a great concern of construction professionals, can be minimized. A number of different green rating systems have been developed to help assess that a building project is designed and built using strategies intended to minimize or eliminate its impact on the environment. In the United States, the widely accepted national standards for sustainable building design are known as the LEED (Leadership in Energy and Environmental Design) Green Building Rating System. The assessment of sustainability using the LEED green rating system is a challenging and time-consuming work due to its complicated process. In effect, the LEED green rating system awards points for satisfying specified green building criteria into five major categories: sustainable sites, water efficiency, energy and atmosphere, materials and resources, and indoor environmental quality; and sustainability of a project is rated by accumulating scores (100 points maximum) from these five major categories. The sustainability rating process could be accelerated and facilitated by using computer technology such as BIM (Building Information Modeling), an innovative new approach to building design, engineering, and construction management that has been widely used in the construction industry. BIM is defined as a model-based technology linked with a database of project information, which can be accessed, manipulated, and retrieved for construction estimating, scheduling, project management, as well as sustainability rating. This paper will present a framework representing the building knowledge contained in the LEED green building criteria. The proposed building knowledge framework will be implemented into a BIM platform (e.g. Autodesk Revit Architecture) in which sustainability rating of a building design can be automatically performed. The development of the automated sustainability rating system and the results of its implementation will be discussed.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권4호
• /
• pp.2171-2195
• /
• 2017

When purchasing an online product, a customer tends to be influenced strongly by its reputation, the aggregation of other customers' ratings on it. The reputation, however, is not always trustable since it can be manipulated easily by attackers who intentionally give unfair ratings to their target products. In this paper, we first address identifying trustable users who tend to give fair ratings to products in online rating systems and then propose a method of computing true reputation of a product by aggregating only those trustable users' ratings. In order to identify the trustable users, we list some candidate features that seem related significantly to the trustworthiness of users and verify the robustness of each of the features through extensive experiments. By finding and exploiting these robust features, we are able to identify trustable users and to compute true reputation effectively and efficiently based on fair ratings of those trustable users.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 2009년도 IWAIT
• /
• pp.703-708
• /
• 2009

The present paper describes the application of an improved impression transfer vector method (Sakurai et al., 2007) to transform the three basic dimensions (Evaluation, Activity, and Potency) of higher-order impression. First, a set of shapes and surface textures of faces was represented by multi-dimensional vectors. Second, the variation among faces was coded in reduced parameters derived by applying principal component analysis. Third, a facial attribute along a given impression dimension was analyzed to select discriminative parameters from among principal components with higher sensitivity to impressions, and obtain an impression transfer vector. Finally, the parametric coordinates were changed by adding or subtracting the impression transfer vector and the image was manipulated so that its facial appearance clearly exhibits the transformed impression. A psychological rating experiment confirmed that the impression transfer vector modulated three dimensions of higher-order impression. We discussed the versatility of the impression transfer vector method.

• 감성과학
• /
• 제10권4호
• /
• pp.523-538
• /
• 2007

본 연구는 의인화된 인터페이스 에이전트와 사용자의 성별이 에이전트에 대한 사용자의 평가에 미치는 효과를 알아보기 위해 실시되었다. 또한 에이전트의 시각적인 사실성의 수준에 관계없이 성별과 정서 표현 방식의 효과가 나타나는지를 알아보고자 하였다. 연구 1은 공감적인 정서를 표현하는 얼굴 사진을 에이전트로 제시하고 에이전트의 성별과 사용자의 성별에 따라 사용자의 평가가 어떻게 달라지는지 알아보았다. 참가자로 하여금 에이전트와 함께 과제에 참가하게 한 다음 에이전트에 대한 느낌을 대인 평가 척도를 이용해 평정하도록 한 결과, 전반적으로 여성 에이전트를 선호하는 경향이 나타났다. 또한 참가자가 이성의 에이전트를 보다 긍정적으로 평가하는 경향을 나타내었다. 연구 2는 시각적인 사실성이 떨어지는 그래픽 캐릭터를 이용하여 에이전트로 제시하고 성별과 정서 표현 방식을 조작하였다. 참가자로 하여금 에이전트와 함께 과제에 참가하게 한 다음 에이전트에 대한 느낌을 평정하도록 한 결과, 연구 1과 마찬가지로 여성 에이전트를 더 긍정적으로 평가하는 경향이 나타났다. 또한 참가자는 참가자의 상태에 호응하는 방식으로 공감적인 정서를 표현하는 에이전트를 에이전트 자신의 정서를 표현하는 조건이나 정서를 표현하지 않는 조건보다 더 긍정적으로 평가하였다. 본 연구의 결과로부터 시각적인 사실성이 떨어지는 에이전트가 사용자와 상호작용할 경우에도 에이전트의 성별의 효과와 공감적인 정서 표현 방식의 효과는 여전히 사용자의 평가에 영향을 미친다는 것을 지지하였다.

• 한국생활과학회지
• /
• 제19권2호
• /
• pp.399-407
• /
• 2010

The purpose of this study was to investigate the visual images of shirts according to coloration and the interval of checked patterns. The experimental stimulus and questionnaires for scaling response were developed for this study. The stimuli were 12 color pictures, in which the gender, coloration(WG: white+gray, WK: white+black, GK: gray+black), and interval(0.5 cm, 1.5 cm, 3.5 cm, 5.5 cm) were manipulated. The 7-point rating scale was used for evaluation of the visual image. Data were obtained from 144 male college students and 144 female college students living in Gyeonggi, Gyeongnam, Seoul, Gwangju, Daegu, and Busan on December 2009. For data analysis, ANOVA and the Duncan-test were used by using the SPSS program. The results of this study were as follows; the visual images for shirts according to the coloration and interval of checked patterns consisted of four dimensions of attractiveness, appeal, youth, and modesty. Gender was the main factor when looking at attractiveness. Coloration had the main effect on appeal, youth, and modesty. Also, there was a significant interaction effect of gender and coloration on attractiveness, while there was a significant interaction effect of coloration and interval on modesty.

• 한국컴퓨터정보학회논문지
• /
• 제19권2호
• /
• pp.201-208
• /
• 2014

본 연구는 조직정치의 개념을 성과평가 행위의 상황적 맥락 요소로 반영한 정치적 성과평가 인식과 구성원 태도 및 팀 과업 특성과의 관계 파악을 목적으로 수행하였다. 195명의 기업 근로자를 대상으로 한 실증연구 결과 개인 편향 및 징벌적 의도에서 이루어진 정치적 성과평가 인식은 조직몰입에 부정적인 영향을, 동기부여적 의도에서의 정치적 성과평가는 조직몰입에 긍정적 영향을 미친다는 사실을 밝혔다. 또한 팀과업 특성으로서의 과업상호의존성은 정치적 성과평가 인식을 증가시킨다는 결과를 도출하였다. 추가적으로 연구결과가 시사하는 바와 향후 연구과제에 대해서도 논의하였다.