The document discusses server consolidation for online games and cloud computing. It provides three key reasons why server consolidation is beneficial for massively multiplayer online role-playing games (MMORPGs): 1) spatial locality in player interactions allows natural workload partitioning, 2) workload is highly variable but predictable, allowing aggregation during off-peak periods, and 3) operators run multiple games, allowing shared infrastructure. It evaluates policies for dynamic zone allocation and finds server consolidation reduces the number of required servers.
The document discusses social network services (SNS) and some of the privacy and security issues that can arise from them. It notes that personal information like real names, education histories, and other details can sometimes be involuntarily leaked through users' profiles and interactions on SNS sites, even if that information is not directly shared, through factors like contact descriptions, images, and relationships. It also discusses how data posted on SNS sites can never be fully deleted from the internet over time.
The document discusses research conducted on involuntary personal information leakage on a Taiwanese social networking site. Through analyzing profile descriptions and relationships between users, the research was able to infer users' real names, ages, schools they attended, and other sensitive information even when users did not explicitly disclose it. The research also looked at how exposed different demographic groups were and surveyed users on their views about privacy on the site.
The document provides guidance on research skills for computer science students. It outlines the key steps in the research process as picking a research area, identifying a problem, writing a paper, submitting to a conference, and presenting the work. It offers advice on each step, such as exploring different research areas and problems, following paper writing best practices, and giving effective presentations through being informative, interesting, and insightful. The overall goal is to equip students with the abilities to independently conduct research and make contributions to their chosen field.
網路安全是一個特殊的研究領域，其中一個原因是在網路安全問題中，"對手"不是文字、影像或任何形式死板板的資料，而是活生生的人；這些製造問題的黑客 (black hat hackers) 終日找尋各種系統及網路漏洞，企圖提出更高明的攻擊方式來獲取各種可能的利益。因此，在網路安全研究中，我們無法"預設"黑客會有什麼樣的攻擊行為，而必須從真正的資料中尋找蛛絲馬跡，從大量資料中發現及解決各種已發生或將發生可能危害使用者資料安全及隱私的行為。在這場研究中，我將介紹 data-driven network security research 並以幾個實際的研究案例來展示真實資料的統計分析可以幫助我們解決什麼樣的安全問題。
Crowdsourcing beyond Mechanical Turk: Building Crowdmining Services for Your ...Sheng-Wei (Kuan-Ta) Chen
?
This document summarizes a presentation about building crowdmining services for data mining research. It discusses using crowdsourcing to generate annotations, evaluate data, and assist with retrieval tasks. Common ways crowdsourcing has been used in data mining include image tagging, data entry, answering questions, and assessing image semantics or photo orientation. However, the presenter argues that crowdsourcing is not just conducting user studies, analyzing user generated content, or using Mechanical Turk. The presentation also describes an implicit crowdmining platform called Pomics that turns photos into comic strips to share experiences.
The document discusses research on quantifying user satisfaction (QoE) in VoIP applications like Skype calls. It presents three key contributions:
1) Developing the first QoE measurement methodology based on analyzing large-scale Skype call data to correlate call duration with network quality factors like jitter and bit rate.
2) Proposing OneClick, a simple framework for crowdsourced QoE experiments based on user clicks to indicate dissatisfaction.
3) Introducing the first crowdsourcable QoE evaluation methodology to verify user judgments.
The document discusses computational social science and three common approaches: macroscopes, virtual labs, and empirical modeling. It provides examples of each approach, including using large-scale Facebook and Twitter data to study language patterns and personality (macroscope), manipulating Facebook data to study emotional contagion and voter turnout (virtual lab), and using Twitter data to predict county-level health outcomes (empirical modeling). Overall, the document outlines how new sources of big data allow computational social science to address challenges of studying social phenomena at large scales.
The document discusses planning and organizing a student organization. It covers measuring the current situation and goals, setting SMART objectives, allocating resources, implementing plans, collecting feedback, and adjusting plans based on results. It also discusses different organizational structures and how to assign tasks based on the structure. The key points are that planning involves understanding the current context, setting clear and measurable goals, developing action plans, monitoring progress, and using feedback to improve. Organizational structures determine how information is communicated and tasks can be assigned through project-based or list-based approaches.
The document provides guidance on research skills for computer science students. It outlines the key steps in the research process as picking a research area, identifying a problem, writing a paper, submitting to a conference, and presenting the work. It offers advice on each step, such as exploring different research areas and problems, following paper writing best practices, and giving effective presentations through being informative, interesting, and insightful. The overall goal is to equip students with the abilities to independently conduct research and make contributions to their chosen field.
網路安全是一個特殊的研究領域，其中一個原因是在網路安全問題中，"對手"不是文字、影像或任何形式死板板的資料，而是活生生的人；這些製造問題的黑客 (black hat hackers) 終日找尋各種系統及網路漏洞，企圖提出更高明的攻擊方式來獲取各種可能的利益。因此，在網路安全研究中，我們無法"預設"黑客會有什麼樣的攻擊行為，而必須從真正的資料中尋找蛛絲馬跡，從大量資料中發現及解決各種已發生或將發生可能危害使用者資料安全及隱私的行為。在這場研究中，我將介紹 data-driven network security research 並以幾個實際的研究案例來展示真實資料的統計分析可以幫助我們解決什麼樣的安全問題。
Crowdsourcing beyond Mechanical Turk: Building Crowdmining Services for Your ...Sheng-Wei (Kuan-Ta) Chen
?
This document summarizes a presentation about building crowdmining services for data mining research. It discusses using crowdsourcing to generate annotations, evaluate data, and assist with retrieval tasks. Common ways crowdsourcing has been used in data mining include image tagging, data entry, answering questions, and assessing image semantics or photo orientation. However, the presenter argues that crowdsourcing is not just conducting user studies, analyzing user generated content, or using Mechanical Turk. The presentation also describes an implicit crowdmining platform called Pomics that turns photos into comic strips to share experiences.
The document discusses research on quantifying user satisfaction (QoE) in VoIP applications like Skype calls. It presents three key contributions:
1) Developing the first QoE measurement methodology based on analyzing large-scale Skype call data to correlate call duration with network quality factors like jitter and bit rate.
2) Proposing OneClick, a simple framework for crowdsourced QoE experiments based on user clicks to indicate dissatisfaction.
3) Introducing the first crowdsourcable QoE evaluation methodology to verify user judgments.
The document discusses computational social science and three common approaches: macroscopes, virtual labs, and empirical modeling. It provides examples of each approach, including using large-scale Facebook and Twitter data to study language patterns and personality (macroscope), manipulating Facebook data to study emotional contagion and voter turnout (virtual lab), and using Twitter data to predict county-level health outcomes (empirical modeling). Overall, the document outlines how new sources of big data allow computational social science to address challenges of studying social phenomena at large scales.
The document discusses planning and organizing a student organization. It covers measuring the current situation and goals, setting SMART objectives, allocating resources, implementing plans, collecting feedback, and adjusting plans based on results. It also discusses different organizational structures and how to assign tasks based on the structure. The key points are that planning involves understanding the current context, setting clear and measurable goals, developing action plans, monitoring progress, and using feedback to improve. Organizational structures determine how information is communicated and tasks can be assigned through project-based or list-based approaches.
This document discusses finding one's passion and offers some suggestions for activities to consider such as joining a university, martial arts club, or parkour club. It notes that with so many choices available, one can feel confused but encourages finding something to do in order to potentially discover a passion.
![20150911 [社团事務] 跑酷社創社審查](https://cdn.slidesharecdn.com/ss_thumbnails/20150911-150912001941-lva1-app6892-thumbnail.jpg?width=560&fit=bounds)
