過去總認為市場是有效率的，投資人是理性的，但近年來，行為財務學的興起，開始證明投資人經常做出不理性的行為，導致投資市場只有少數人為贏家，能長期獲得超額報酬。散戶的投資行為也是學術上時常探討的，而在過去幾年，國外開始有學者利用社群網站twitter，並用情緒辨識軟體將發文者的文章做情緒的歸類，來看能不能提升對大盤或個股的預測能力。結論是其中幾類情緒的確能提升預測大盤或個股的機率。而在台灣，比較有名的股票社群為PTT的股票板，該板通常會同時有1000人以上在線上。本文做了兩項研究，其一為藉由本週看多及看空文章的總數來看是否能預測未來幾週的大盤報酬，其二為藉由個股看多或看空文章的回文數，來看是否看多或看空文章的回文數愈多，是否能預測個股未來幾天的報酬，而以上兩項的結果，即使有些部分在統計上為顯著，但並沒有一致性，舉例來說，有些個股可能在第三天顯著，並與回文數呈負相關，但另一些個股可能在第5天顯著，並與回文數呈正相關，因此並無法得到一致性，很難有合理的解釋，因此可說明，PTT裡的發文及回文對整個投資市場來說，並無資訊內涵，並無法預測股市的報酬。 / The efficient market was popular before and most of the investors are seen as rationality. However, behavioral finance emerged in recent years. This theory tells us investors usually invest irrationally that there are only a few people in the market are winner in the long run. Retail investors’ behaviors are usually discussed in theses in recent years. In foreign country, there were some people studying the social network- twitter and use some software to judge the different emotions in an article in twitter. Finally, they found some types of emotions could predict the movement of stock prices or stock index like Dow Jones. In Taiwan, the stock section in PTT is famous. Many people express their opinions in stock section of PTT. In this paper, author use the bullish or bearish article to forecast the stock index a few weeks later and use the number of responses of an article for a specific stock to predict the return of the stock a few days later. The results are that some of the responses of stocks can predict stocks returns five days later and there are negative relation between the numbers of responses and the returns. However, other responses of stocks can predict stocks returns three days later and there are positive relation between the numbers of responses and the returns. The author can’t find the consistent results. There are more than1000 people stay in the stock section of PTT at the same time, but these people are only a small group in the stock market. Consequently, we can use PTT to predict the stock market.

近年來，越來越多的社群網站受到人們廣泛的使用，例如：Facebook、Plurk之類的網站都擁有大量的使用者資料。社群網路越來越受到一些研究學者的重視，很多人開始紛紛研究如何有效的運用社群網路上的資料。影響力的傳播是社群網路上一個很重要的課題，如何在社群網路上找到影響力最大的組合是個受到廣泛討論的問題。在本研究中，我們試想一間公司如果要請人來宣傳產品的話，必須支付代言人一些費用，而如何在有限的預算下聘請一些代言人來達到最大的宣傳效果就是我們研究的問題。兩個代言人的影響力總和並不單單只是將兩個代言人的影響力相加而已，因為代言人本身所影響的對象可能會重複，所以必須扣除掉一些被重複影響的人，也增加了問題的困難度。在我們提出的演算法中，可以有效的減少計算的時間並且使找出來的代言人組合所造成的影響力趨近最佳解。 / Influence maximization is the problem of finding a small subset of nodes (seed nodes) in a social network that could maximize the spread of influence. In this paper, we given a social network and budget, which people should we choose could maximize the spread of influence with a limit budget. We propose a new algorithm combine cluster algorithm and dynamic programming to solve this problem. Our experimental results show that our propose algorithm achieves better running time comparing with the CELF algorithm. But CELF algorithm achieve much better influence spread than our propose algorithm. Based on our results, we believe if we can improve the cluster algorithm than we can achieve much better influence spread.

Social networking for enterprises is the use of emergent social software platforms within or between companies to enable participants to rendezvous, connect or collaborate through computer-mediated communication and to form online communities (McAfee, 2006). It provides a platform for participants to co-create social networks that permit the exchange and retrieval of relevant knowledge on a company-built platform. The objective of this research is to understand the characteristics and development of social network applications in different types of enterprises. We discuss the key factors for companies to develop their own social network using web 2.0. Based on the classification of customizing products (Pine, 1993), this study uses multiple case studies and the characteristics of their social networks such as service, marketing and collaboration to see how social networking on web 2.0 helps customize products for business benefits. The findings of the study are that with different types of customizing products, the company will have a different focus in its value chain. We explored these differences by constructing a two-way model. Using the two-way model, our study separates the different types of products into four sectors: provider, listener, promoter and collaborator. The four sectors can describe the purpose of the product using the social network and gives advice that is relevant to similar types of companies.

隨著個人電腦與網際網路科技的逐漸成熟，網路上每日都有巨量資料（Big Data）產生。近年來隨著社群網站的崛起，如何處理這些巨量的社群資料，並有效率地提供出有意義的社群資訊，將是這幾年社群網路領域研究的重點。每當內閣改組消息一出的時候，各政府部門單位的官職繼任官員，都會成為社會公眾關注的議題。本研究將使用中華民國政府官職資料庫，以社群網路分析與連結預測理論為基礎，並透過資料庫中所提供的資料，隨著不同評選時間點以及評選官職建置出網路。擷取網路的資訊，利用本文所提出的多面向模型（Multiperspective Model）產生多種觀點的分數。接著使用評選模型（Evaluation Model）將各個觀點的分數整合，進行某官員繼任某官職可能性計算，然後輸出官職繼任官員的評選清單（Evaluation List）。最後對輸出的評選清單分別對空降繼任狀況、各級上司對於繼任人選決定影響力、單一觀點與多觀點評選方式的評選結果、多觀點評選方式下重視的觀點，以及官職繼任成因五項分析進行探討。 / With the well development of personal PC and the Internet technology, there is a huge amount of data (Big Data) being generated on the Internet every day. Because of the debut and rise of social websites, how to deal with such a huge amount of community information as well as efficiently provide meaningful data to the public has been an explored main issue in the field of social network research in recent years. When the news about cabinet changing was released, the successor of various government departments will become the actively concerned topic for the public. This research applied a government position transaction database as the elements to build the network, which based on Social Network Analysis and Link Prediction theory with different evaluation position and evaluation time. Captured information in the network was used to generate the scores of multiple perspectives according to the Multiperspective Model. Then using the Evaluation Model, which can integrate each observed perspective, and calculate the probability of an official succeeds of a position. Finally the network could output the evaluation list of position successor. In the end, the outcome of the evaluation list was applied to analyze and discuss the following 5 research questions: The situation that the successor isn’t from the unit of successive position, the influence of all levels superiors on the succession decision, result of evaluative methods of a single view and multiple perspective, the important perspective of Multiperspective evaluation, and causal relationship of official successor.

隨著部落格的文章越來多，如何在大量的部落格文章中有效的幫助使用者找到需要的文章就變成一個非常重要的問題。因此，本研究結合部落格搜尋和個人化搜尋的技術，提出針對搜尋部落格文章的個人化部落格搜尋。其中，主要包含了兩種個人化部落格搜尋的方法。首先，延伸一般個人化搜尋的方式針對部落格的環境加以修正，提出了內文個人化部落格搜尋。透過分析和搜尋系統使用者相關的部落格文章內文，讓使用者可以更快的在部落格中找到需要的文章。接著，我們提出創新的社交個人化部落格搜尋，藉由分析使用者和其他部落客之間的社交行為，來達到個人化的目的。最後，在實驗中可以發現，個人化部落格搜尋明顯的提升了部落格搜尋的準確度和滿意度。 / As the quantity of articles increased rapidly, how to help the users to find the blog articles effectively becomes an important issue. Therefore, this study combines the traditional blog search and personalized search techniques, and proposes a way to search blog by personal information. This study mainly contains two methods of personalized blog search. First, the general way of personalized search environment for blogs to be amended, the content based personalized blog search. By analyzing and searching the system-related blog article text of the users, this study can make users quickly figure out the articles they need. We propose a social based personalized blog search, by analyzing the social behavior between users and other bloggers, to achieve personal goals. Finally, the result of the experiments has shown that both ways obviously improved search accuracy and satisfaction.

由於近年社群網路各種應用網站興起，像是Facebook、Twitter，等，相關議 題也逐漸受到討論，例如越來越多利用社群網路傳播訊息或者病毒式行銷的相關 研究。當我們能夠找出一個社群網路當中，習慣的傳播模式或者是傳播路徑，並 且能從中定位各種角色的重要性，進一步在社群網路中找出這些角色後，在這些 相關的議題的應用將更加靈活。 目前各大社群網路應用網站，使用者都可以與社群網路中的好友分享自己的 動作，例如發佈影片或圖片，評論，按「讚」等，基於這樣的前提使用者的任何 活動是有機會被社群網路中的好友影響，因此我們定義好友間影響的可能性，以 及依觀察合理的定義出社群網路中較為重要的角色。 我們的演算法經由收集使用者在固定社群網路應用網站的各種動作，加上動 作的時間所形成的動作誌(action log)，以及使用者們所構成的社群網路，可以從 社群網路中找出主要的資訊傳遞路徑以及各種不同限制下的領袖以及追隨者，並 且將會利用社群網路應用網站驗證分析我們所定義的角色成為結論。 / Recently, social networking services and websites such as Facebook and Twitter are taking more and more parts in our daily life. Issues of in uence propagation have been studied in recent years. To ll in the gap of previous works, we aim to discover the main path of in uence and dene the importance of leader in hierarchy on the social graph. Social networking users are in uenced by the power of social networking service as they are able to post and likevideos, pictures and comments. Therefore, in this study we propose to discover the possibility of a relation and important roles by mining social activities. After collecting performed action and time stamp from dierent users and understanding their social network, our framework was able to identify the main in uence paths and leaders under dierent constrains. Most importantly, our approach outperforms both on precision/recall and ranking in realistic data.

社群網站是目前蓬勃發展的網路應用服務，其多樣化功能與高互動性不僅影響使用者生活習慣也促進其購買行為。越來越多企業投入大量成本經營社群網站，希望從中得到好處，但仍有部分企業質疑經營社群無法替企業帶來實質效益。本研究目的在探討經營社群是否可以帶給企業效益。藉由大量文獻探討，提出了六個關於社群效益的假設。透過四階段研究來驗證假設，第一階段蒐集了全球16個跨國企業資料加以分析，得到了一個概略性的了解; 第二階段蒐集了58個跨國企業的資料進行社群能力與產品銷售之關聯性分析;第三階段研究將焦點放回台灣企業，蒐集了電信、航空、銀行以及便利商店四個產業中11個企業資料進行社群能力與產品銷售的關聯性分析;第四階研究段針對第三階段的11個企業進行社群經營與企業效益的分析。 本研究發現，經營社群與提升顧客滿意度以及提高品牌知名度上關聯性較低；而在航空，銀行以及便利商店產業中，經營社群可以提高整體收益。此外，企業提供具有1)較低價位、2)較低耐受度的產品或服務以及3)消費者參與度較低的購買決策較容易受到社群經營影響。更進一步發現，企業以提升顧客成長以及強化顧客忠誠度為目標去經營社群會獲得較高的成功率，也較容易得到實質效益。 / Most of the major enterprises have invested in social networks to provide services, promote products, or communicate with customers. However, to date no clear understanding on the effects of the social networks on business performance has emerged. This study examined the relationship between the capability of the enterprise to manage social networks and the resultant benefits of their investing in the social network. A fan page on Facebook is the selected target of study. By analyzing the efforts required to manage a fan page, we measure the fan page management capability by its reach, referrals, activities, and interactions. By analyzing literature and practical cases, we hypothesize six types of business benefit impacts can be experienced by enterprises investing in a fan page. The study applied four-stages of data collection and testing. The first stage is to build a general understanding of the relationship between the number of fans and business revenue. Based on findings of the first stage, the study tested the fan page benefit hypotheses on selected global firms and sought patterns of benefits generated from the fan page. In the third and fourth stage, the study tested the fan page effect on product sold of selected firms in Taiwan and further examined other hypotheses by industries. This study builds deep understanding from various aspects about fan page effects on enterprises, and proposes that firms provide products and services with characteristics of low-product price, short-product durability and low-customer involvement are more influenced by the management of their fan pages. Retail industries such as: banking and airline with high capability of managing Facebook fan page are more likely to increase customer growth and intensify customer loyalty.

隨著社群媒體的普及，新聞媒體與意見領袖逐漸重視在社群媒體上以貼文方式發佈新聞資訊，社群媒體成為許多使用者會接收新聞與重大事件的主要管道且透過社群媒體的評論、分享與按讚等互動機制表達立場，這些即時互動行為是傳統媒體缺乏的機制，如何分析也是研究上的挑戰。本研究將針對Facebook上的貼文與互動行為進行分析，提供一款互動視覺化系統，找出貼文資料集中相似的貼文群集以及隨著時間推移下貼文屬性的變化，進一步瞭解Facebook上使用者、貼文與重大事件之間的相互影響。由於Facebook上的貼文與互動行為具多維度屬性，我們透過降維演算法將大量的貼文以二維散佈圖呈現，達到將相似貼文分群的效果。另外，我們設計了一種視覺化呈現方法，「Time Block」，突顯出時間的推移下貼文屬性的變化，藉此觀察出貼文資料集是否存在特定的模式。最後提供即時互動的操作介面，以及貼文屬性以及關鍵字兩者的統計，藉此連接到貼文集的屬性與時間的分佈關係，協助以視覺化方式進行探索與分析。最後，透過案例分析與使用者測試呈現此視覺化探索工具的優缺點。 / Social media becomes an essential medium for broadcast news. News media and option leader post information and people love to receive news and interactive using comment and likes to feedback. It is a research challenge to analysis this massive amount interactive behavior data in social media. In this paper, we propose an interactive visualization system to explore on the posts and interactions on Facebook. This system can help a user to find out the similar interactive behavior cluster and the trend of time-varying attributes to understand how the users, posts, and a big event to affect each other. Facebook Posts and interactive behavior contains multiple dimensional attributes; we adopt the dimensional reduce algorithm and 2D scatter plot to present the cluster in the spatial domain. Then, we design a time-varying visualization method, `Time Block,' can highlight the changing attributes and observe the unique pattern in the time domain. Also, we design a real-time interactive interface to connect the cluster and trend visualization with additional keyword distribution and attribute statistics. Finally, we use case study and user study to demonstrate the advantage of the proposed system.

在耐延遲網路環境下節點的相遇情況不是很頻繁，這可能導致節點間的連線斷斷續續，使得有效地將訊息傳遞成為一件困難的事情。藉由社群感知轉送機制的中間度指標特性，可以來提升傳送成功率。雖然大多數研究幾乎都是將訊息轉送到單一目的地或是多個且已知的目的地。然而，一些應用像是廣告的散佈，要將訊息送給對訊息有興趣的人，但卻不知道是誰。因此，關鍵的問題為如何建立社群網路關係的親密度機制，來選擇作為轉送訊息的節點，並利用群體廣播的方式盡可能有效地傳播至最多可能目標目的地，進而提升效能。 本論文以群播機制和社群感知當作基礎概念，來設計新的轉送訊息的方法和公式化選擇中繼節點的機制。最後，我們使用政治大學實際軌跡來模擬，將模擬結果與其它路由演算法比較，其結果證明我們所提出的方法能提高訊息傳送成功率和正確率，降低傳送延遲時間和傳送訊息的成本。 / In delay tolerant networks (DTNs), nodes infrequently encounter with each others. This results in intermittent connectivity of the nodes, and makes it difficult to deliver the message effectively. A social-aware forwarding scheme can help for successful delivery ratio by utilizing the characteristic of their centrality metric. Most of the previous studies focus on message delivery to single destination or some priori known destinations. However, some applications like advertisement dissemination may not know who will be the interested persons to be delivered. Therefore, the key challenge is how to establish the social relationship strategy to select appropriate nodes as relays, and furthermore to use multicasting to disseminate effectively as many “target” destinations as possible to improve the performance. This thesis developed a new strategy which has a new forwarding message scheme and formulates the selection of the relay nodes based on the concept of the multicasting and the social network. Finally, we used the reality trace data of National Chengchi University to simulate. The simulation results are compared to others DTNs routing protocols as well as other social-aware forwarding schemes. The results showed that our proposed approach can enhance the successful delivery ratio and delivery accuracy, decrease the delivery delay and reduce the delivery overhead.

網路購物興盛、社群網站湧現以及資訊搜尋的方便性，持續性資訊搜尋成為購買決策很重要的影響因素。本研究先蒐集相關文獻與彙整搜尋動機的元素，共包括九種資訊搜尋動機，分別是：探索動機、滿足動機、價值動機、思想動機、產品提供豐富度、產品資訊提供、方便性、角色動機以及社交動機，並將這九種元素歸納為三種資訊搜尋動機，分別是：享樂性、功利性以及社交性，再探討搜尋動機與持續性社群網路資訊搜尋對購買意圖的影響關係。本研究透過網路發放問卷，有效樣本共480份，採用量化分析，並以最小平方法(PLS)進行有效樣本之實證分析。 本研究經因素分析發現，探索動機與滿足動機同屬於一種構面，本研究將此構面命名為：「探索滿足」；產品提供豐富度、產品資訊提供、方便性同屬於一種構面，本研究將此構面命名為：「資訊可得」；角色動機與社交動機同屬於一種構面，本研究將此構面命名為：「社交需求」。此外，在思想動機與價值動機方面，本研究則是分別重新將這兩種構面命名為：「資訊潮流」與「優惠需求」，以更符合理論邏輯與這兩種因素的意涵。 本研究結果顯示：研究分析歸納的五種資訊搜尋動機：探索滿足、資訊潮流、資訊可得、優惠需求、社交需求，對於持續性社群網路資訊搜尋皆具有正向的影響；持續性社群網路資訊搜尋對購買意圖亦具有正向的影響關係，而產品需求強度對於持續性社群網路資訊搜尋與購買意圖之間的關係則是不具有調節效果。這代表無論消費者對於產品是否有需求，只要平時在社群網站上瀏覽到產品的相關資訊時，就能影響人們購買意圖的產生，這間接證實了非計劃購買發生的可能，也突顯了持續性資訊搜尋的重要性。最後，本研究結果期望可以提供個人或業者於社群網站上的產品銷售策略之經營與學術未來研究之方向。 / Due to the flourishing of Internet shopping, the emergence of social network sites and the convenience of information search, ongoing search has become a very important purchase decision factors. In this study, we collected relevant literature and aggregated a total of nine kinds of search motivations, namely: adventure, gratification, value, idea, product offerings, product information, convenience, role and social. The nine search motivations are summarized into three search motivations, namely: hedonic, utilitarian and sociality. Then, we explore the effect relationship for the search motivations and ongoing social network information search to the purchase intentions. Then, we did questionnaires via the internet and a total of 480 valid samples were received, after that, we used quantitative analysis and used partial least square (PLS) to conduct empirical analysis. By factor analysis, we found that adventure motivation and gratification motivation belonged to the same factor, we named “exploration meet”. Product offerings, product information and convenience motivation belonged to the same factor, we named “information availability”. Role motivation and social motivation belonged to the same factor, we named “social needs”. In addition, we re-named idea motivation to “information trend” and value motivation to “preferential needs” to make these two kinds of factors be more consistent with logical and theoretical implications. The result of this study showed that the five kinds of search motivations: exploration meet, information trend, information availability, preferential needs and social needs all had positive impacts on ongoing social information search. Ongoing social information also had positive impacts on purchase intentions. But the product needs intensity didn’t have a moderating effect between ongoing social information search and purchase intentions, this result not only indirectly confirmed possibility of unplanned purchase, but also highlighted the importance of ongoing search. Finally, the results of this study could be expected to provide marketing strategy for persons or industries on the social network sites and could be referred for the future of academic research.