WORK IN PROGRESS
This is an outline for Phase 1 (24 months) of a Bamboo Implementation Proposal.
The purpose of this document is to provide information to institutions and organizations participating in the Bamboo Planning Process so that they can help determine (1) the long term future of Bamboo and (2) define what activities Bamboo will carry out in its first implementation phase. The intent of this document is to solicit community input toward the ongoing development and revision of the implementation proposal. As this is an early draft, it is not yet a commitment to carry out all or any of this work.
Please note that we are updating this document frequently based on wide ranging input from the Bamboo community. These updates will occur periodically and will be indicated as ".1", ".2", ".3", etc updates. In addition, we will occasionally make major document revisions. These are noted as "1.X", "2.X", and so forth. Between major document revisions there may be some inconsistencies in language used between the sections of the document.
4.2 - Bamboo Atlas
The Bamboo Atlas will enable articulation, collection, review, and discovery of scholarly methods and digital technology that supports scholarship. Functionality - storage, organization, annotation, rating, review, search, and discovery - will be optimized to support the formation of communities of interest in humanities scholarship and support.
Scholars, librarians, and technologists will each find avenues into a rich store of practices and generalized workflows (a.k.a. "recipes") derived from those practices. Generalized workflows will include hyperlinked references to digital tools; to digital content; and to defined activities that involve using tools to examine, organize, annotate, and transform content. Tools will be further linked to technology services from which they are partly or wholly composed. Each of these elements and the relationships between them may be richly annotated with reviews, ratings, and descriptive tags (keywords). Atlas consumers will be able to use multiple methods of search and discovery that align to different user perspectives. Materials of interest and relevance can be surfaced via tailored filtering based on reviews, ratings, and tags related to self-identified areas of interest. Alternatively, filters may be based on contributions by trusted members of a scholarly community.
Initial scholar participation will be seeded by humanities faculty, including leaders in their discipline, whose participation will form a part of their institution's contribution to Project Bamboo. Incentives for ongoing contribution to the Atlas is key to the effort's success; incentives will include peer-reviewed publication of scholarly methodologies linked to Project Bamboo, and Atlas participation as an element of partnership with tool, application, and content projects (described in the Bamboo Services Platform section of this proposal). The Atlas will be further seeded by integration with existing efforts to catalog practice and technology as applied to the humanities. Over time, a gravity of proven utility and desire to participate in a vital network of engaged, collaborative scholarship will draw more and ongoing participation.
The Bamboo Atlas will be built as a services "back-end" that will enable the initial interfaces developed through close, iterative consultation with faculty and other Atlas users. The Atlas services will support a proliferation of interfaces, including unanticipated "mashups" implemented by interested parties outside or at the periphery of the Bamboo community. Human and machine interfaces to the Atlas materials will be enabled to permit presentations and analyses of the data by general or specialized communities of interest, as well as by those interested in humanist methodologies and their intersections with technology as an object of scholarship in and of itself.
18.104.22.168 Atlas elements and function
The technology services that realize the Bamboo Atlas will focus on the following principal elements:
- Elements of Scholarly Methodology
- Scholarly Narratives - Description of particular aspects of scholarship, scholarly methodology, research, and/or teaching from the scholar's point of view, and in her own language.
- Recipes (generalized workflows) - Recipes describe how to achieve goals using information technology. Recipes are written for scholars and use non-technical language to describe the tools, digital content, and steps (activities) needed to complete work. In short Recipes generalize the particulars articulated in Scholarly Narratives into activities that occur across multiple processes of scholarship, and enable librarians and academic support professionals to develop technology to support those activities.
- Activities - Steps (units of work, process, or procedure) that occur in the course of scholarship.
- Tools - Software that a scholar uses to accomplish her goals. A tool usually aggregates and organizes a number of related software capabilities for the user's convenience.
- Content - Digitized objects of scholarship. These may be digitized texts, images, audio, video, etc.; and may themselves be artifacts of prior scholarship.
- Elements of Shared Technology Services
- Services - A unit of software that delivers a related set of capabilities (functionality). Services implement capabilities that have been decomposed then logically grouped to flexibly facilitate (a) interoperability with other software and/or digital content; and/or (b) combination and recombination with other services in support of multiple tasks or workflows. A tool may aggregate, organize, and present the capabilities of multiple services.
- Capabilities - In a service, a piece of work (software functionality) that the service can deliver on request.
- Service Family - A group of related services.
The Bamboo Atlas will enable the following interactions between a community of scholarship and Atlas elements:
- Contribution - Authoring (solo or collaborative) of each Atlas element, and assertion of relationships between them (e.g., Recipe X is a generalization of Narratives Y and Z)
- Categorization - Atlas elements may be categorized in an individual's or in group-curated collections; and collections may be shared with individuals, groups, or the world
- Annotation - Atlas elements and the relationships asserted between them may be annotated with "tags" (i.e., associated with a word or phrase from a curated or an unconstrained vocabulary)
- Review - Participants may contribute a review of any Atlas element, including collections of elements
- Rating - Participants may rate any element or relationship between elements on multiple scales (e.g., interestingness, utility, degree of generality/specialization, etc.)
- Discussion - Participants may comment on Atlas elements, relationships, and metadata in "threaded" discussions
- Community - Participants will be able to organize themselves in communities of self-defined interest, sharing profile information that enriches opportunities to connect
- Search - Participants may search based on key words; occurrence of specified text in an Atlas element; association of an element with one or more tags; ratings; existence and character of reviews associated with Atlas elements; etc.
- Discovery - Participants may browse Atlas elements filtered by association with a community of interest, trusted person, disciplinary topic, content type, language, annotation/tag, etc.; and will benefit further from unanticipated discovery services whose invention will be enabled by the open programming interface (API) to Atlas materials.
Back-end (behind the scenes) functionality will:
- harvest and link to catalogs or registries of method, practice, tools, and services maintained by other projects and communities, enriching direct contributions by Atlas users with broad and deep effort that already supports humanities scholarship; and,
- track usage of the Atlas, and present the information collected in ways that enrich search and discovery functionality
- support notification to users who want to follow the evolution of Atlas elements, including metadata associated with and relationships between them.
A variety of user-interfaces will be enabled by services that provide access to Bamboo Atlas information; only a subset of those interfaces will be built by Bamboo partners. However, the range of that variety will be broadened or narrowed by service design decisions made by Bamboo partners. An iterative process of user-centered design will be employed to maximize the breadth of possibility. This process will include consideration of precedent and best practice in areas of Atlas function and interface, and exploration of how the Atlas can be tailored to specific contexts of humanities scholarship.
Here are some examples of existing precedent and practice in user interface design and function - mostly from well known commercial sites - that might both illustrate the concepts described in this proposal, and inform aspects of Atlas function and interface:
Search, Discovery, Trust
- Amazon.com includes a book's citations and citations by other books, (see, for example, R. Ellmann, Yeats: The Man and the Masks, (Norton, 2000)).
- Amazon.com includes key phrases "mined" using multiple algorithms from a book's text, (see, for example, R. Ellmann, Yeats: The Man and the Masks, (Norton, 2000)).
- Amazon.com includes a concordance of the most frequently-occurring words in a book, as well as additional statistics mined from a book's text (see, for example, the concordance for W. G. Sebald's Austerlitz (Modern Library, 2002))
- Amazon.com suggests similar items in its catalog based on usage patterns (in Amazon's case this is based on other books bought by those who bought the book being presented).
- Flickr.com is a well-known example of how "tags" can function as a kind of one-word summary of the content of an image - tag visualizations on Flickr can search as a discovery mechanism, and can be focused by popularity or recent contribution.
- Flickr.com uses geotagging to allow access to its billions of images by place
- Flickr.com combines multiple sets of information, such as "where the clickthroughs are coming from; who comments on it and when; who marks it as a favorite; its tags and many more things which are constantly changing" to arrive at a measure of "interestingness," which serves as a path for discovery based on what others have found valuable.
- In a (scientific) research context, Nanohub.org presents recommendations based on other items viewed by those who viewed the item being presented (see, for example, the Nanohub entry for a Crystal Viewer tool)
- An example of "faceted search" - the ability to search not just by keyword, but by categories of information - can be seen in the Flamenco list of Nobel Prize winners
- Wikipedia.org exemplifies how deep hyperlinking - affording ability to wander idiosyncratically through a body of information, finding one's serendipitous way - can offer rich opportunities for discovery (start, for example, at the disambiguation page for the term Austerlitz, which can leads one to a novel referenced in examples in this list, as well as through articles on several cities, on the eponymous Paris train station, and on the performer Fred Astaire).
Review and Rating
- Amazon.com includes views of a smaller, more prominently presented set of reviews, and a larger more complete set (see, for example, the initial, smaller and more detailed & complete presentation for W. G. Sebald's Austerlitz (Modern Library, 2002))
- Amazon.com presents contributed reviews and ratings of a book in aggregated (chart) form and in detailed modes that permit viewers to see how other viewers judged the reviews/ratings, e.g., "most helpful" (see, for example, reviews and ratings for W. G. Sebald's Austerlitz (Modern Library, 2002))
- Facebook.com allows a simple expression of approval via its one-click "Like" function that can be applied to the status updates, photos, and other content contributed by one's social network ("friends" in Facebook's jargon)
- Apple Computer's support forums allow someone who submitted a question to mark a response "helpful" or "solved," which awards "points" to the person who responded. A participant's accrued points boost the level of a status marker that appears next to each of the participant's contributions, affording viewers a quick way to judge how the community ranks or trust contributions from active participants.
Present information tailored to different interests and audiences
- Wikipedia.org presents articles on a topic in different languages. Some of these articles may be or may originate in translations of a single contribution, while others are presentations of a topic contextualized by contributors whose perspectives are situated in different national or linguistic frames (see, for example, the English-language article on Gare d'Austerlitz, the Paris railway station; this article is translated or differently-presented in fourteen languages as of 25 May 2009).
- Dell Computers is notorious for organizing its website to obscure content by forcing users to choose a perspective (such as "Home" or "Public Sector") that influences - most often narrowing - the range of products presented as available for purchase. The Bamboo Atlas will want to avoid that trap.
- Many Eyes is a data visualization lab run by IBM. From tag-cloud examples to maps portraying where the "worst drivers" can be found in the U.S., this site gives some sense of the breadth of possibility when a well-defined service back end opens data to visualization "mashups."
- Amazon.com's concordance (also used as an example in the Search, Discovery, Trust section, above) displays the most frequently-occurring words in a book as a tag cloud (see, as above, the concordance for W. G. Sebald's Austerlitz (Modern Library, 2002))
22.214.171.124 Atlas-enabled activity
Visitors and contributors to the Bamboo Atlas will participate in a number of meta-activities, some simply in the course of using the Atlas, and others as deliberate explorations and analyses of Atlas content and usage. These meta-activities include:
- User-centered design: Bamboo members will work together to observe how users interact with the Atlas in order to refine its functionality and interface.
- Contribution as a mode of scholarly communication: Contribution to the Bamboo Atlas will be linked to avenues of scholarly communication, such as publication in digital journals focused on methods of scholarship in the arts, humanities, and interpretive social sciences.
- Design and review of technology services: The Atlas will serve as a forum to express the nature and methodology of humanist practice, and to respond to evolving implementation of technology services that attempt to support scholarship.
- Evaluation of scholarly practice: Scholars will use the collection of information in the Atlas to consider the nature and evolution of scholarship in the arts, humanities, and interpretive social sciences.
- Linking research and pedagogy: Research practice and method represented in the Atlas will be available as source material for pedagogy in modes of humanities scholarship; and pedagogy represented in the Atlas will be available as source material for research in pedagogy.
- Assessment of the impact of Project Bamboo: Patterns and shifts in modes of scholarship will be traceable through usage activity collected about services offered and/or tracked by the Bamboo Atlas. These trends can inform assessment of Bamboo's qualitative and quantitative impact on its community.
126.96.36.199 Incentives to Atlas contribution
Integration of effective incentives to participation is fundamental to the success of the Bamboo Atlas. Incentives might include particular elements or functionality in the Atlas, and might be modeled on successful community-building functionality in "Web 2.0" sites, platforms, and other contexts (e.g., Amazon, Flickr, Wikipedia - see sidebar). Incentives might also include linkage of other activity, such as deployment of services on the Bamboo Services Platform or publication of methodology in peer-reviewed journals, to Atlas contribution. Efforts to identify incentives have begun in the Bamboo Planning Phase, and will continue as a vital strand of Implementation Phase work.
Incentives proposed to date include publication of Atlas contributions in a digital journal (e.g., Atlas contribution of narratives and recipes documenting humanist methodology as a path from or to a dedicated, peer-reviewed vehicle of scholarly communication). It is expected that partnerships described in the Bamboo Services Platform section of this proposal will include an element of contribution to the Bamboo Atlas as well.
Incentive suggestions are being actively solicited for inclusion in this proposal. Please contact David, Chad, or any of the Program Staff with your suggestions.
The Bamboo Atlas will connect the ebb and flow of teaching and research with the constant changes of technology. Shifting ideas about how practices, resources, and services can and should be linked together will be implicit in the Atlas, and available for discovery, analysis, and interpretation by humanists with more and less engagement with technology-enabled methodologies. Dialogue generated in the course of exploration and review will connect scholars with technologists, researchers with content providers, students with instructors and practitioners, and domain specialists with engineers.
Awareness: The Atlas will function as a living library to record, preserve, and make available information that traces the developing course of digital scholarship in the humanities. While it cannot be complete, it will leverage user contributions as well as harvested and linked information from catalogs and registries maintained by others. In doing so, the Atlas will offer tailored access to descriptions of practice and method, and provide pointers to methods, tools, and digital content applicable to the interests of Atlas users. Ability to dynamically view and "mine" (use software to detect patterns in) evolving resources and analysis, and draw connections to broader contexts and categorizations, will help faculty, students, librarians, funders, institutional leaders, technical architects, and service developers to engage in and support humanities scholarship.
Insert here: a humanist's take on how Awareness will affect her work (multiple, different perspectives - particularly from faculty with different degrees of engagement with technology-enabled methodology - would fit well here)
Connection: Shared profile information and the ability to form communities of interest around contribution, review, rating, and annotation of Atlas content will foster connections between scholars across disciplinary, institutional, and international borders.
Insert here: a humanist's take on how Connection will affect her work
Collaborative Design: Scholars' narratives and "recipes" (generalized workflows) will inform technologists of humanists' real-world needs, while ratings and reviews provide direct feedback that will shape revision and evolution of technology services.
Insert here: a humanist's, technologist's, librarian's, or iSchool faculty member's take on how Collaborative Design will affect her work
Consideration of practice: Means to surface and assess usage patterns and shifts in modes of research and pedagogy will inform scholarly assessment of technology's impact on the humanities, as well as assessment by funders and campus stakeholders.
Insert here: a humanist's or iSchool faculty member's take on how Consideration of practice will affect her work
4.2.4 Work Plan
Realizing the Bamboo Atlas will involve multiple strands of effort that will strongly influence each other in iterative increments. Some effort can (and must) occur in parallel, with appropriate coordination points and processes. Implementation Phase efforts will be informed by Proof of Concept accomplishments during the Planning Phase. At a high level, elements of Implementation Phase strands of effort are:
- Year One
- Refine concepts and attributes of Atlas artifacts (Narratives, Recipes, etc.) and the processes by which they are collected, analyzed, and annotated
- Develop and realize incentives to participation in Atlas activity
- Model an initial base set of user interfaces to Atlas functionality, using an iterative user-centered design process
- Collect, contribute, analyze, organize, annotate, review, and rate Atlas content, both during and after service development phases
- Model services API and data structures to support Atlas functionality, conforming to relevant standards and taking into account the above in iterative stages of model development
- Build services to support Atlas functionality
- Identify valuable external sources of information to harvest and/or link to from the Bamboo Atlas
- Build software that will appropriately ingest information from sources other than participant input
- Deploy Atlas services on a Bamboo Services Platform
- Build an initial base set of user interfaces to Atlas functionality
- Deploy an initial base set of user interfaces to Atlas functionality
- Year Two
- Model, design, implement, and deploy enriched functionality and interfaces based on experience and analysis of Atlas usage and value
- Model, design, implement, and deploy functionality to enable community-curation of vocabularies (taxonomies, ontologies, synonymies) to enrich search and discovery mechanisms
- Model, design or adapt, implement, and deploy service registry functionality
Each of the elements listed will have many component pieces of work; these are fleshed out further in the Detailed Work Plans section of this proposal. The high level effort for Year 1 (2010) might be represented graphically, in sequential layers and mutually-influencing strands:
- Productionized Bamboo Atlas services
- Productionized base set of user interfaces
- Body of Atlas materials (contributed and harvested)
- Incentive program(s) to attract participation
- Services Registry model or adoption-plan
- Year 1 Atlas entities will include Narratives, Recipes, Activity Definitions, Tool References, Content References; but will not include Services Registry elements.
- Year 1 Atlas metadata entities will include Discussions, Reviews, Ratings, and Annotations (Tags).
- Year 1 community entities will include People, Profiles, and Community Groups.
- Year 1 Atlas services will enable all interactions listed in 4.2.1, above, at a useful and usable level of sophistication
- Year 1 Atlas services will expose administrator-curated vocabularies (designed and refined by Bamboo Plan partners)
- By the end of Year 1, an extant Services Registry implementation will be selected, or a model for a Bamboo Services Registry will be sketched. The intention is to implement/adopt Services Registry functionality in Year 2.
- Enriched user-profile
- Enriched search and discovery functionality
- Enrichment and evolution of Year 1 Atlas functionality (based on user-centered design feedback)
- Community-curated vocabularies (taxonomies, ontologies, synonymies)
- Services Registry
As the Bamboo Community sharpens its conception of the Atlas, modeling and design complexities will arise in both predictable and unanticipated areas. Some of the non-trivial issues likely to require resolution include:
- What should be versioned? Only major entities (e.g., narratives, recipes)? Reviews and ratings? Relationships between tags and entities?
- Should the relationships asserted between major entities be "version aware"? What (if anything) should happen when a relationship has been asserted between Narrative X and Recipe Y, and Y is revised? Must the assertion be revisited? May it be optionally re-asserted vis-a-vis the new revision of Y?
- Should superceded (old versions of) content be indexed, or should search mechanisms only index current versions of things like narratives, recipes, reviews, and ratings?
- If multiple groups form, and the full membership of each decides that the groups ought to be merged, how should the Atlas track pre-merge group activity (such as group-contributed reviews or ratings, assuming such activity is enabled)?
- If a scholar uses a certain word to "tag" Atlas content, then decides she would rather use a more broadly-employed equivalent term, should her original choice disappear, or should it remain part of the Atlas in some inactive or less-preferred form?
Collaborative, Editorial, and Curatorial Activity
- Is all collaborative contribution a collaboration among peers, or might a role of principal author, curator, owner, or editor-in-chief apply to certain modes of engagement with the Atlas?
- Should an editor/curator be able to assert "this is a variant of artifact X" and release that variant to the world, allowing the original author to actively accept, actively reject, or ignore the proposed variant? Or should the submission be to the original author, allowing her to decide which suggestions to accept and reject?
- Should a contributor be enabled to grant a "collaborator" or "editor" role to whomever she wishes? Should such a role be grantable for each contribution, for all contributions, for contributions of certain types (e.g., all recipes, but no narratives), or all the above?
- Should an individual or group be enabled to keep a contribution private until the contribution is judged ready to share? In such cases, should the contribution's evolution (versions) be made public, or only the ready-for-release draft - or should both options be available?
- How should the Atlas incorporate or refer to material maintained by other tools, sites, or archives?
- Copy/cache or simply refer?
- What if a referral URL/URI is not permanent?
- Should there be ongoing, automated link-checking; and, if so, what ought to happen when a link is discovered to be broken?
- What metadata could be identified and stored to facilitate finding a cached copy of an artifact (e.g., in the Internet Archive) should originally-referenced material go away?
- What arrangements can and should the Bamboo Atlas make with copyright holders of materials that are image-digitized but for which OCR'd text is not directly available? E.g., in the case of journal articles about humanist methods that are held by JSTOR, should the Atlas be designed to federate search across the JSTOR index (cf. XML Gateway) on demand (when a user explicitly searches) ... or should the Atlas attempt to incorporate JSTOR search results into its own map of humanist methodology?
- How can information of value to the Bamboo Atlas be identified and appropriately culled or referenced out of sprawling, unfocused archives in which it may reside, e.g., in the HATHI Trust collections, or even on the web as a whole?
Privacy vs. Value
- How can the Atlas be designed to encourage sharing of contributions that enrich search and discovery functionality?
- Should Atlas contributors be permitted to maintain private collections of information on Atlas infrastructure?
- Private to themselves alone, to groups they belong to, to institutions with which they are affiliated?
- Is there a type of obfuscated sharing that ought to be minimally required - such as profile attributes that are publicly exposed without association to individual identity; and public concordances of words that occur in contributed material associated with identity-obscured profile information but not with the private material itself?
- Should sharing information with internal Bamboo Atlas indexing functions be required? What about sharing with analytical software that uses the Bamboo Atlas as a source of information, but resides outside the "boundaries" of data governed by access permissions that Atlas users define?
- Should a participant be allowed to dissociate her Atlas-profile from public-domain publications, associations, and other expressions of ideas that no longer fit her point-of-view? Or should such information be preserved, with a provision for the participant to declare her current relationship to it (e.g., "I no longer believe this is true"?
- Which elements of the Atlas should be maintained in alternate languages for internationalization purposes, if any?
- Which translated elements ought to be maintained as part of the core Atlas information, and which as part of an internationalization service "layer" atop the services that provide an interface to core Atlas information?
- How should the Atlas services be modeled to facilitate declaration of the language in which a submission (narrative, recipe, review, tag) is being contributed? Should the Atlas record whether submissions contain text in multiple languages, and if so how? (Note a begged question here, which is an independent category of modeling challenges: how and to what degree and using what standards should textual contributions be marked up with respect to semantic content, formal structure, and/or presentational format.)
- How can Atlas services be modeled to best interact with services that machine-translate textual content?
- How can Bamboo best engage partners with expertise in language issues, such as CLARIN?