京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第18563号 / 工博第3924号 / 新制||工||1603(附属図書館) / 31463 / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 藤田 正治, 教授 中川 一, 准教授 竹林 洋史 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第20344号 / 工博第4281号 / 新制||工||1663(附属図書館) / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 白土 博通, 教授 八木 知己, 教授 KIM Chul-Woo, 教授 杉浦 邦征 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第20346号 / 工博第4283号 / 新制||工||1663(附属図書館) / 京都大学大学院工学研究科都市社会工学専攻 / (主査)教授 清野 純史, 教授 高橋 良和, 准教授 古川 愛子 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

No description available.

Fake news has great potential to cause damage to brand reputations and finances. Given the technical challenges of detecting fake news in time, it is inevitable that social media platforms will end up hosting fake news. The competition for attention and advertising revenue is intense. Many consumers read only the headlines. Fake news stories that mention brands in headlines can help news publishers garner social media engagement but can also hurt brands, raising concerns about brand protection. In this research, I focus on the first two stages of the information processing model – attending to information and encoding information (Berk 2018; Miller 1988).In Chapter 2, I investigate whether mentions of human and product brands are associated with news consumption and news retransmission (how brand mentions attract attention; the first stage of information processing). Using data from a news platform that generated both traditional and satirical (fake) news stories, I quantify the effects of brand mentions on social media engagement for both traditional and fake news. The analysis encompasses mentions of popular product brands, such as Apple, and mentions of human brands, such as famous politicians and actors. A framework based on uses and gratifications theory (UGT) aids in variable selection and the interpretation of results. My results imply that human brand mentions generally have a positive effect on news consumption and retransmission for both news formats, and product brand mentions affect engagement of satirical news via an interaction with news categories. Results provide further insights on the roles of sentiment, narrative style, and writing quality of news stories. The high potential of human and product brands in the headlines, especially human brands in satirical news, may be indicative of their potential to be misused by unscrupulous news media publishers. This reminds social media platforms of their responsibility to protect brands and consumers from fake news. Next, in Chapter 3, I examine the effectiveness of before-warnings (BWs) and after-warnings (AWs) in alerting consumers and reducing the persuasive influence of fake news on brand attitudes (how warning timing affects encoding; the second stage of information processing). Results reveal that for both negative and positive fake news, BWs are sometimes no more effective than no-warnings. Although BWs do encourage more critical processing of misinformation, this can distract consumers from the warning message. More importantly, Chapter 3 demonstrates a robust after-warning effect (AWE). Warning consumers after they have read fake news with AWs consistently leads to a higher reduction of persuasive influence (negative or positive) than BWs. AWs are more salient and arouse greater reactance to the false information than BWs. The resulting loss in control over how the news influenced attitudes and increased anger lead to the observed after-warning effect. News valence also matters since positive news is perceived to be more credible and processed less critically than negative news. AWs relative to BWs thus arouse feelings of being tricked when fake news is positive but not when it is negative, also leading to the after-warning effect. The findings have several theoretical and managerial implications. / Business Administration/Marketing

An association was tested between the presence of a television weather broadcaster on-screen and viewers’ likelihood to seek shelter, measured via risk perception and preventative behavior. Social networking websites were used to recruit respondents. Four clips of archived severe weather videos, one pair (on-screen and off-screen broadcaster) using the reflectivity product and another pair (on-screen and off-screen broadcaster) using velocity product, were presented to participants. Viewers’ trust and weather salience were also quantified for additional interactions. A relationship between viewers’ risk perception (preflectivity = 0.821, pvelocity = 0.625) and preventative behavior (preflectivity = 0.217, pvelocity = 0.236) and the presence of the broadcaster on-screen was not found. The reflectivity product was associated with higher risk perception and preventative behavior scores than the velocity product (prp = 0.000, ppb = 0.000).

A new safety application, as part of ITS Advanced Rural Transportation System (ARTS), has been developed to be deployed on a two-lane rural road (Route 114), in Southwest Virginia. The route segment under study is subject to significant head-on accidents, as a result of two main conditions: 1- Illegal passing maneuvers crossing solid yellow line, and 2- A short passing sight distance due to the road vertical profile. The main objective of this research is to design a video detection-based warning system by installing an affordable and efficient system on the vertical crest curve on Route 114, capable of performing the following two main functions: 1.Detect vehicles that attempt to violate the no-passing zone restriction (i.e. when crossing into the opposing direction). 2.Warn the drivers violating the restriction in order to discourage them from continuing their maneuvers. System architecture as well as detailed system design was developed. A system simulation was conducted with the use of a special software program written with MATLAB. The simulation was applied for both "with" and "without" the system cases. The simulation runs showed that the system could virtually eliminate all head-on collisions, should violators obey the early warning messages displayed. Several sensitivity tests were made for different scenarios. Finally, the viability of the system was evaluated from economic point of view. The financial analysis revealed high economic indicators. / Ph. D.

The radio frequency spectrum is becoming increasingly crowded and research efforts are being made both from the side of communication and from radar to allow for sharing of the radio frequency spectrum. In this thesis, suitable methods for classifying incoming signals as either communication signals or radar signals using machine learning are evaluated, with the purpose of filtering communication signals in radar warning receiver systems. To this end, a dataset of simulated communication and radar signals is generated for evaluation. The methods are evaluated in terms of both accuracy and computational complexity since both of these aspects are critical in a radar warning receiver setting. The results show that a deep learning model can be designed to outperform expert feature-based models in terms of accuracy, as has previously been confirmed in other fields. In terms of computational complexity, however, they are vastly outperformed by a model based on ensemble decision trees. As such, a deep learning model may be too complex for the task of filtering communication signals from radar signals in a radar warning receiver setting. The classification accuracy needs to be weighed against the model size and classification time. Future work should focus on optimizing the feature extraction implementation for a more fair classification time comparison, as well as evaluating the models on recorded data. / Radiospektrumet blir alltmer belastat och forskningsinsatser görs inom både kommunikation och radar för att tillåta delning av spektrumet. I denna rapport utvärderas lämpliga metoder för att klassificera inkommande signaler som antingen kommunikation eller radar med hjälp av maskininlärning, med syftet att filtrera ut kommunikationssignaler i radarvarnare. För detta ändamål genereras ett dataset med simulerade kommunikations- och radarsignaler för att jämföra modellerna. Metoderna utvärderas med avseende på både precision och beräkningskomplexitet, eftersom att båda aspekterna är kritiska egenskaper i en radarvarnare. Resultaten visar att en djupinlärningsmodell kan utformas för att överträffa modeller baserade på expertdesignade särdrag med avseende på träffsäkerhet, såsom tidigare visats inom andra områden. Avseende beräkningskomplexitet, är däremot modellen baserad på en ensemble av beslutsträd överlägsen. Detta innebär möjligen att en djupinlärningsmodell är allt för komplex för syftet att filtrera bort kommunikationssignaler från radarsignaler i en radarvarnare. Modellens träffsäkerhet bör vägas mot dess storlek och tiden för klassificering. Framtida arbete bör inriktas på att optimera beräkningen av särdragen för en mer rättvis jämförelse av tiden som krävs för klassificering, samt att utvärdera modellerna på inspelad data.

This study investigated the effects of cross-modality warning presentation and retention in a dual-task paradigm in a simulated military environment under various task-induced stress levels. It was also intended to determine what role working memory played in the mode of warning presentation that resulted in the highest retention and subsequent compliance. An all within participant design was created in order to determine if scores on working memory span tasks predicted performance across the varying forms of warning presentation. Furthermore, task-induced stress levels were varied over the course of the experiment to identify if workload transitions affected performance. Results revealed that when the presentation format and the response format matched (e.g., verbal-verbal), behavioral compliance was greater then when presentation and response format were mismatched (e.g., verbal-pictorial). Thus, it is not necessarily the presentation type that affects compliance, but the combination of presentation and response mode. Analysis also revealed that the pictorial-pictorial warning combination resulted in greater behavioral compliance compared to verbal-verbal or written-written combinations. The format of warning presentation did not affect performance on the operational tasks as predicted. Thus, the visual/spatial operational task, regardless of its complexity was not interrupted in timesharing with intra-modal warning presentations or cross-modal time-sharing. As predicted, task based stress affected the WCCOM task in all experimental procedures. Results further revealed that as task demand increased, performance on the WCCOM task decreased. Task demand did affect the operational tasks, the shooting and the navigation tasks. The shooting task, which was less complex than the navigation task was not affected by lower levels of task demand, but at the greatest level of demand (eight warnings) performance in the operational task, degraded. Degradations in performance on the more complex task, the navigation task, materialized at a moderate level of task demand (four warnings). For subjective ratings, task demand did affect workload ratings. As the task demand increased, the subjective workload ratings also increased, revealing a true association between workload and subjective ratings. The working memory separability hypothesis was supported by the working memory span tasks, but consequently they were not predictive of the warning presentation format.

Project schedules should mirror the project, as the project takes place. Accurate project schedules, when updated and revised, reflect the actual progress of construction as performed in the field. Various methods for monitoring progress of construction are successful in their representation of actual construction as it takes place. Progress monitoring techniques clearly identify when we are behind schedule, yet it is less obvious to recognize when we are going to slip behind schedule. This research explores how schedule performance measurement mechanisms are used to recognize construction projects that may potentially slip behind schedule, as well as what type of early warning they provide in order to take corrective action. Such early warning systems help prevent situations where the contractor and/or owner are in denial for a number of months that a possible catastrophe of a project is going to finish on time. This research develops the intellectual framework for schedule control systems, based on a review of control systems in the construction industry. The framework forms the foundation for the development of a schedule control technique for forecasting schedule slippage — the Required Performance Method (RPM). The RPM forecasts the required performance needed for timely project completion, and is based on the contractor's ability to expand future work. The RPM is a paradigm shift from control based on scheduled completion date to control based on required performance. This shift enables forecasts to express concern in terms that are more tangible. Furthermore, the shift represents a focus on what needs to be done to achieve a target completion date, as opposed to the traditional focus on what has been done. The RPM is demonstrated through a case study, revealing its ability to forecast impending schedule slippage. / Master of Science