The first part of the thesis focuses on Information Foraging Theory which was developed by Peter Pirolli, a cognitive scientist from Intelligent Systems Lab at Palo Alto Research Center, to understand how human search in an information environment (Pirolli 1995). The theory builds upon the optimal foraging theory in behavioral ecology, which assumes that people adapt and optimize their information seeking behavior to maximize the success of accomplishing the task goals by selectively choosing paths based on the expected utility from the information cues. The expected utility in Information Foraging Theory is called Information Scent. The second part is to design and build a new way to visualize search engine results in a graphical way that incorporates the concept of information scent to make search experience more efficient for users. The end result of the project will be an improved visualization of search results, obtained by using Google’s Application programming interface (API), latent semantic analysis, and data visualization methods to present a semantics-based visualization of the search results. The proposed design is developed to increase information scent for relevant results and shorten the foraging path to reach the search goal by presenting users with fewer but more valuable proximal cues, thus making search a more human-centered experience.

Major research studies have provided support for information scent based usability evaluation and have increased its parent theory's (Information Foraging theory) credibility in the HCI community. These studies have, directly and indirectly, found significant correlations between good information scent and good usability. We would like to investigate its application to less-studied platforms, such as web pages on PDAs and cell phones. The theory itself is not device specific and it implicitly assumes that information scent's importance is universal. However, all studies on the practical application of Information Foraging theory have been conducted with desktop computers. We would like to examine what role information scent plays in interfaces on mobile devices that are limited in usable screen space. For this project, we performed a controlled study with 28 participants on the BBC News web site and its PDA-optimized version. Various usability and information scent related indicators were measured and compared across devices. Contrary to our expectations, we did not find any statistically significant differences between the information scent indicators of the Desktop and PDA sessions and the paths across the devices were highly correlated. / Master of Science

This dissertation sought to explain goal achievement at limited traffic “long tail” Web sites using Information Foraging Theory (IFT). The central thesis of IFT is that individuals are driven by a metaphorical sense of smell that guides them through patches of information in their environment. An information patch is an area of the search environment with similar information. Information scent is the driving force behind why a person makes a navigational selection amongst a group of competing options. As foragers are assumed to be rational, scent is a mechanism by which to reduce search costs by increasing the accuracy on which option leads to the information of value. IFT was originally developed to be used in a “production rule” environment, where a user would perform an action when the conditions of a rule were met. However, the use of IFT in clickstream research required conceptualizing the ideas of information scent and patches in a non-production rule environment. To meet such an end this dissertation asked three research questions regarding (1) how to learn information patches, (2) how to learn trails of scent, and finally (3) how to combine both concepts to create a Clickstream Model of Information Foraging (CMIF). The learning of patches and trails were accomplished by using contrast sets, which distinguished between individuals who achieved a goal or not. A user- and site-centric version of the CMIF, which extended and operationalized IFT, presented and evaluated hypotheses. The user-centric version had four hypotheses and examined product purchasing behavior from panel data, whereas the site-centric version had nine hypotheses and predicted contact form submission using data from a Web hosting company. In general, the results show that patches and trails exist on several Web sites, and the majority of hypotheses were supported in each version of the CMIF. This dissertation contributed to the literature by providing a theoretically-grounded model which tested and extended IFT; introducing a methodology for learning patches and trails; detailing a methodology for preprocessing clickstream data for long tail Web sites; and focusing on traditionally under-studied long tail Web sites.

Internet portals are increasingly becoming a primary source of information. A portal is a gateway to information on the Internet or a hub from which users may locate relevant information (Strauss, 2000). Because university Web sites have various user classes, universities are beginning to adopt the portal concept for their Web sites. The study conducted aimed to determine the effect of tailoring information content and presentation style on a university Web portal. User ratings of information design on three metrics and user task performance measures of time and errors were compared for four prototypes. Three prototypes were built on the basis of user requirements and two Human-computer Interaction (HCI) theories and one was a replica of an existing academic information portal. The three metrics were derived from the HCI theories. The contributions of the study are a determination of user acceptance of and user performance with the tailored presentation styles and three metrics derived from HCI theories that can be used to compare alternative information presentation styles for portals. An important contribution is the remote data collection technique that was used in the study and a time-stamping technique that recorded clicks on hyperlinks. / Master of Science

Think-aloud protocol is a usability testing method whereby the participant running the usability test on an interface, thinks aloud as a way of giving feedback of the task he/she is performing on the given interface. It is one of the most researched on usability testing methods. It has attracted both praises and criticisms based on the effects it has on the participants or the tests at hand. A recently done study that used simple tasks, aimed at finding out the difference between using think-aloud protocol and not using think-aloud protocol. The study concluded that no notable differences were evident on the number of fixations and the amount of screen areas viewed when using think-aloud protocol and when not using think-aloud protocol.As an extension and follow-up of the recently done study, this study focused on finding the type of tasks that the concurrent think-aloud protocol has effects on. The tasks were chosen based on the information scent concept and eye-tracking methodology was used in collecting the necessary results.The study that involved twenty participants, resulted to some effects of the concurrent think-aloud protocol being noted on the low-scent tasks but not on high-scent tasks. It therefore goes ahead to conclude the tasks onto which concurrent think-aloud protocol would be more effective and the tasks that would be executed more effectively through other usability testing methods other than concurrent think-aloud protocol.

面對大量習慣用嘗試錯誤方式去學習使用複雜線上遊戲功能的使用者，如果沒有一個好的網頁佈局設計，會妨礙讓使用者易於上手。本研究試圖綜合以往在電腦學習認知方面的模板比對理論、CoLiDeS Model、及資訊氣味等相關研究，增強對線上遊戲介面設計的了解。本研究設計一個線上遊戲介面認知的實驗，用「仙劍Online」多人線上遊戲的介面為實驗工具，運用眼動儀去觀察受試者的瀏覽行為。「功能的群組位置」(Location)與圖像(Icon)皆為介面設計的重要因素；這兩項因素可能會影響受試者對任務難易的知覺判斷與瀏覽行為的Pattern。這些差異會顯示在任務的「間隔多久時間看到正確區域位置」、「凝視次數」、「凝視正確區域佔全部的比例」、任務的「完成率」與任務的「使用時間」等資料上。經過本研究驗證可以明確的知道「功能的群組位置」(Location)與圖像(Icon)何者為設計者首要考量之因素，在學術上能加強對CoLiDeS Model的了解，提出兩階段修正的認知模式；在實務上幫助介面設計師進行更有效的頁面設計和實用的建議。 / A large amount of users practice the trial and error approach to learn to use complicated online games. This leads to low user perception of "easy to use" when the game lacks good webpage layout design. This study attempts to integrate previous studies on cognitive learning through computer including Template Matching, Information Scent, and Comprehension-based Linked model of Deliberate Search Model ( CoLiDes Model), to enhance the understanding on online game interface design. A cognitive experiment of online game interface was designed utilizing the multiplayer simulation game "Chinese Paladin Online" in an attempt to observe and record user’s browsing behaviors through the eye tracker. "Location of functional group " and "Icon" are important factor of interface design and affect user’s browse behavior pattern . These differences will appear in "Time spent", "Accuracy", "Frequency of fixation ", “ and "Time spent before first fixation on the correct position" of the tasks. We expect the results extended the CoLiDeS Model, enhanced understanding of the interface layout user cognition, and provided fruitful suggestions in how to further design on effective WebPages.