CSCL Research Support: Semantic Knowledge Tags, Tabs, and Lenses

Jason Lee

May 5, 2008

INFO 608

CSCL Research Support: Semantic Knowledge Tags, Tabs, and Lenses

A Conceptual Design Paper

Fundamental to any research effort, including CSCL research, is the process of gathering knowledge from various sources, categorizing that knowledge in a way that is useful when retrieving or sharing that knowledge, storing that knowledge, and ultimately retrieving and communicating that knowledge. Many mechanisms are either proposed or in existence that will aid the CSCL researcher in performing these basic functions to research. However, advances in global connectivity (Web 1.0), interactive Web technology (Web 2.0), and more recently Web-based data management (Web 3.0) provide opportunities to reexamine the possibilities of exploiting these advances for the purpose of advancing the art of research in general – including CSCL research.

Knowledge Tagging

Imagine a researcher is currently writing an article on a new method of conducting CSCL using the latest in Web technology. The researcher searches online articles or publications for information on the current research between CSCL and that Web technology and discovers a publication with experimentation results which reinforces why CSCL might greatly benefit from using that Web technology. The researcher “tags” that information as pertinent to the term “CSCL”. In addition the researcher adds the technology name to the list of possible terms, links that term to CSCL, and then tags that new information with the new tag as well. The result is the researcher has tagged the information with both tags and the tags also explicitly linked.

Advances in Data and Knowledge Representation

The World Wide Web(Web 1.0) itself was born out of a need to way for physicists to share information(History of the World Wide Web, 2008). It obviously had immediate and explosive popularity and transformed itself through the needs of many users into a relatively easy-to-use network of static Web sites. The transformation continued and the Web became more dynamic to the point where the Web site content and even the overall user experience of a Web site could changed based on who was browsing and what they were doing. This transformation is generally known as Web 2.0 and is the inspiration for some of the concept in this conceptual design paper. However, there continues to be a transformation on the Web in the way data and relationships among data are captured and made available. This is known as the Semantic Web or even sometimes as Web 3.0. In the Semantic Web, data about data (i.e., metadata) is stored in an ontology that defines the data elements related to a particular topic the relationships of that data to other data elements in other topics. These data elements and topics are fully-qualified such that all ambiguity of terms is eliminated and links between ontologies can be created, creating a Semantic Web (Semantic Web, 2008). The Web was born out of the research community and the research community should continue to benefit from the advancements in Web concepts.

Knowledge Tagging using Semantic Concepts

The system is designed exploit the conceptual ideas of the advancement of Web interactivity and the Semantic Web. As described in the scenario, the system will allow the researcher to categorize information found from various sources using a categorization scheme that is both common to the research field and also personal to the researcher. Since the collection of information that is found is the key element of this functionality, this portion of the systemshould be product-based. In other words, the user interface should be centered on manipulating the product (i.e., information) and not necessarily the process involved in manipulating the product. Since the researcher will be adding information to a personal and shred repository, categorizing the information, and in short manipulating the information, the most appropriate interaction type is the manipulating interaction type. Users will be able to drag and drop tags from a “box” of all the tags or just a subset of favorite tags or commonly used tags onto information on the users’ screens. Likewise, tags can be removed from information. Information can be collected and remain untagged for an indefinite amount of time and can be tagged at any time. Further explanation of the conceptual design for the interface is explained in a later section called “Knowledge Tabs”.

The interface metaphor of “tagging” is used since is a term that is gaining popularity on the Web currently and evokes a feeling of placing a tag on something so that describes that thing. Essentially what we want the researchers using this system to feel the same way about the information they are “tagging”. This metaphor is familiar to the researcher community and provides a familiar structure to the research community and to the Web user community in general. Much like a user of a physical tag would be able to denote information on the tag, choose from a variety of tags, and place the tags on pertinent items, the user of this system will be able to do the same but just to non-physical information and using non-physical tags. Additionally, the tagging interface metaphor is relevant to the problem of trying to categorize information since that the primary purpose of the physical tag is categorization. The interface metaphor is easy to represent in a visual manner. The user of the system will be able to visualize the tagging of the information in both a tab format, which is explained in a later section as a different interface metaphor, and inline as an iconic representation of the tags related to the information. More information about the tag, continuing the metaphor, can be viewed by manipulating the tag (perhaps by hovering over the tag). The metaphor is also extensible, such as removing or replacing a tag on an object, or could be combined with other metaphors easily.

Collaborative Expansion of the Knowledge Tags

When a baby is born, a tag is given to the baby – a blue tag for a boy and pink tag for a girl. This is a tagging system that is common throughout much of the world. Imagine if someone wanted to change the system such that a baby boy over 10 pounds is denoted with a dark blue tag. Would this change be useful? Perhaps to some people in some situations. Perhaps it would be critical to visualize this information for other. How would one go about making this change? It would be a long process of acceptance for such a simple change to be implemented through the entire world of neonatal care, if it happened at all.

However, in regard to our problem of information categorization, a change to the information “tagging” system (e.g., collection of ontologies in the Semantic Web) could be easily changed and adopted across a wide range of users. This could be accomplished in the system since each researcher is concerned with a different range of information that overlaps to a certain degree with other researchers. As one researcher changes the tagging system by adding new tags or associating tags together with connectors or verbs others will be able to use those changes immediately by tagging their information with the new tags. Additionally, the researchers will be able to see the new implicit relationship among their tagged information and the new information which has been linked due to the new tag association added to the tagging system. New avenues of research could be opened up due to the new associations being made by other researchers.

Knowledge Tabs

Another interface metaphor that could be applied to this system is one of “tabs”. This metaphor comes from the use of tabbed file folders to store paper documents containing information. When a researcher associated information to a particular tag, that information can be viewed in a “tab”. The user can decide which tags are to be viewed as tabs. The information associated with that tag are displayed for the user in that tab, including information that is associated with a subset of that tag. Also information could be tagged with multiple tags. So the user will be able to click on other tags associated with that information. If that tag has a tab available, that tab will become active. If not, the tab will be created, loaded with the associated information, and made the active tab. Users can also browse other tags for pertinent information by clicking on the tags in the tag “box” (explained earlier). The tag will become a tab and the user can browse the information and follow the semantic links if desired. However, the user will only be able to edit information in the user’s personal area. However, the information can be dragged into the user’s personal area from another user’s area and edited. Of course, all the associated tags from the copied information (including citations for references) are copied over to the personal area as well. The combination of the tag and tab interface metaphors allows the user to organize and view the information in two familiar and interconnected ways. Likewise, the combination of the manipulating and exploring interaction types enables the user to work in the user’s area and continually expand the user’s knowledge of related research and incorporate that information into the user’s area.

The “tab” interface metaphor provides a structure that is understandable and familiar to the general user. The metaphor is related to the problem of categorizing, storing, and retrieving information. However, the interface metaphor will need to be extended to include clicking on information in one tab and activating (or opening) another tab. However, this system assumes some level of familiarity with Web interaction. The metaphor is easy to represent with the exception of the physical storage of tabbed folders in a filing cabinet. Since a tabbed folder could conceivably hold anything object, the metaphor could be extend to hold any type of information in our system (i.e., multimedia).

Translation Lens

There is a potential language barrier between researchers that are working in the same or overlapping aspects of CSCL. As a user is browsing information from another user’s area and that information is in another language that the user does not understand, the user must translate the information to make it useful. Additionally, if the user copies the information into the user’s area, the information should be translated.

The interface metaphor of a lens can be used to promote the user experience of looking at information through a lens so that it becomes focused and understandable. The user does not need to be aware that the information being viewed originated in another language – just as someone using eyeglasses does not need to be aware of the contrast between the blurry image that would be evident without the eyeglasses and the sharp image seen by using the eyeglasses. Of course, one can remove or put on eyeglasses at will. Similarly, the user of the system should be able to “remove” or “put on” the Translation Lens at will through a configuration performed using the user interface. Continuing the interface metaphor, one could only need eyeglasses sometimes (i.e., reading) and does not need to use them on all the time. Similarly, the user of the system could turn on the Translation Lens for a piece of information or a tab at a time through a configuration of the user interface. This would be particularly helpful when a user is fluent is a few languages that are frequently used in the user’s field of study and does not want the information translated into one language, which could potentially lose some contextual meaning or translate idioms incorrectly.

It might not be a good idea to translate the tags. That could get messy since one-word translations without context are often incorrect. However, one of the features of the Semantic Web is the ability to create synonyms. As users find tags that are pertinent, they can add synonyms to the knowledge tag repository for tags that are the exactly the same but are in different languages.

The interface metaphor of a lens provides structure in that the user will generally be familiar with how eyeglasses are used or will understand the metaphor. When the Translation Lens is activated there is a high degree that the user will imagine looking through a lens at the screen. The only consideration to note is that the user might forget the Translation Lens is activated and leave it on when it is not desired. The user interface implementation would have to sufficiently signal that the Translation Lens is activated. However, the lens interface metaphor is highly relevant to the problem of translating text from one language to another, though it is not the only metaphor that could be imagined to solve the problem. The metaphor is easy to represent through an iconic representation of eyeglasses and the fact that the user will be looking at a screen (e.g., lens) at the information. The metaphor is extensible – the use of varying eyeglass prescriptions could be used as a metaphor for how much should be translated or which languages should be translated.

Summary

The conceptual design for the system explained in this paper exploits the advancement of the Web, in particular the Semantic Web. Information is crucial to the any researcher, including a CSCL researcher. Tagging that information and linking that information together is also crucial when trying to conduct research. Using the interface metaphors of tags, tabs, and lenses this system will help CSCL researchers collect, categorize, and share information.

References

Semantic Web – Wikipedia. (2008). Retrieved May 4, 2008, from Wikipedia Web site:

History of the World Wide Web – Wikipedia. (2008). Retrieved May 4, 2008, from Wikipedia Web site:

Page 1 of 5