Traditional approaches to content management have evolved from glorified FTP Servers (File Transfer) Protocol) or combinations of file systems with overlaid databases controlling access in a more elegant manner—but with the cost of higher system complexity and resource overhead. As these systems transformed from document management to web content management, the degree of overlap across their respective technologies became evident.

Most content management systems continue to be built on the idea that all users are aware, and proficient in, the way the system stores and manipulates information. This mindset is really not too far from the file and folder structures embraced wholeheartedly at the end of the last decade.

As the advent of Web 2.0 indicates, the dogmatic notion that users must be on intimate terms with the inner workings of a content management systems is flawed. Realising that Roundbox Global has created the RoundboxCMS™ Learning Object Repository (LOR) that distills content assets and learning objects to an atomic or granular level.We refer to this as Molecular Content Management™. As a key component complementing tools in an organisation’s learning technology portfolio, RoundboxCMS™ enables content, asset and object reuse without complex conversion, import or export. Molecular Content Management™ is the concept of establishing asset relationships similar to the way molecular bonding creates elements. Content and learning object contributors require little knowledge of the content universe stored or managed within RoundboxCMS.Assets distilled to a granular level can be assembled to deal with learning outcomes, learner remediation and ancillary topics that support learning objectives. A granular ability to address content enables tracking assets for rights and use management, while offering the larger community the ability to access content objects that do not need to be created locally. For example, a math object could be created once and reused across geographies.

In addition, semantic analysis – a technique derived from natural language processing – is used to analyse the relationship between documents and the terms they contain.

Unlike static data such as document date and time of creation, semantic analysis actually delves through the text to come up with statistical breakdowns about the nature of the text. In this manner, the RoundBoxCMS™ engine helps:

• Identify key concepts and their significance within the context of the overall information environment
• Determine how concepts are related to one another
• Create and maintain dynamic concept maps
• Determine which content is unique and what is duplicated
• Understand relations between documents, content and assets
• Use embedded information in documents (such as links), to further refine understanding

Designed to mimic human reading and analysis, the engine “understands” a document’s semantic relevance by identifying and analyzing the concepts and their relationships via innovative data representations, set algorithms, and heuristics. Just as Google’s Page Rank algorithm has been inspired by human heuristic of annotated references (links), the Latent Semantic Engine uses comparable insightful heuristics to improve document analysis and search. The engine’s humanistic relevance metric provides a more intuitive ranking than simple statistical measures. Molecular Content Management doesn’t require manual tagging, which means that data can be processed, sorted and made ready for use with less human intervention.

The coupling of Molecular Content Management™ and LSA alleviates the burden of content organisation by automating tasks previously needing human effort. Structuring information differently can generate new knowledge and concept maps, providing intelligence tools that help analyze soft trends such as user interest, etc.