Untitled
Segue > Crescendo (04:55 mp3)
Crescendo is being envisioned as a knowledge base management system. It should include such functionality as uploading, sharing, searching, filtering, tagging and summarizing knowledge bases and collections. A knowledge base here refers to a given user’s collection of knowledge assets as well as their relationship to an existing knowledge collection.
Uploading Knowledge Collections
Users will need an interface for uploading their knowledge collections (e.g. iTunes library.xml files). This upload UI should have options for limiting upload to a subset of their collection (e.g. upload all records where genre = Language-Japanese). Uploaded files should then be parsed and each record should be inserted or updated in the database.
In addition, users should be add records directly to the knowledge base. If a user’s role is one of a student, their record is tagged as student and added to their own knowledge base. Knowledge collections managers should also be able to batch upload language learning assets, merge new data into existing assets and add student records into a knowledge collection.
Searching Collections
Once a given user’s collection is in the database, they should be able to search the metatdata of own collection much like they can in iTunes. The search UI should include summative statistics that provide a profile of their collection (e.g. average rating/knowledge of a given lesson)
Those records in a given user’s metadata collection that match a language learning asset should include all parts of that asset that the user is authorized to view (e.g. Japanese students should be able to play audio for Japanese language learning assets).
Tagging Collections
For any given asset a user has access to, they should be able to add both structured and unstructured tags. Structured tags are tags specified for a given collection (e.g. knowledge tags) that can be either quantitative or qualitative (e.g. knowledge tags with attributes ranging from 1-100 or with attributes good, bad, ugly). In addition, the application code should be able to generate tags in certain modes of usages (e.g. knowledge tags should be added by the application based on user responses in an assessment mode)
The attributes of structured tags should be determined and weighted by application context. Certain application contexts (e.g. assessment mode) should result in higher quantitative attributes than other contexts (e.g. self evaluation).
Tracking Collections
Any given user should be able to view all the records that match a structured tag in their own collection (e.g. view all records with the tag knowledge whose attribute in the range of 1-50). As well, they should be able to view all the records that match an unstructured tag in both their own and the collections of other users (e.g. view all records that have been given the tag difficult)
Collection managers should be able view the structured tags of language learning assets that any and all users who have access to their collection have specified, as well as generate summative statistics (e.g. average knowledge for all assets in lesson one of the Genki Japanese collection)