Present: Amit Kumar, Matt Kirschenbaum, Bill Parod, Catherine Plaisant, Stan Ruecker

Stan reported on the work of the Interface cell. They are going ahead with developing an "Eclipse like" web based environment or workbench that will help with the construction of modular and pluggable-in units. There will be a conceptual evaluation by late July and some proof-of-concept implementation by Labor Day. Discussion focused on the question whether this takes us away from more immediate tasks and whether or how development can be folded into SEASR, a project that is more about a development framework than a particular project.

We talked about 'monkeyfying' FeatureLens, which currently is a standalone program.

Discussion about the data cell and the model of a data store confirmed that we are working for a while on a dual track model, with exploring the capacities and scaling problems of the Nora and WordHoard data models. Some 250 English novels from 1780-1900 (~40 million words) have now been tokenized and morphosyntactically tagged with MorphAdorner, and will provide some immediate opportunities for testing scale issues.

There was some disccusion, crossing various cells, about what kinds of activities to log and at what level of detail. The "system" keeps track of what it does, and logs of various kinds are useful for figuring out what gets done, how well or fast different procedures perform, not to speak of storing, repeating, or sharing particular queries. This topic clearly will require more discussion across different cells.

Matt reported about discussions in the Collaboration cell about a 'publication model.' This will involve 'playing and replaying' Monk operations.

Martin reported on some conversations with Tim Cole at UIUC and Matt Jockers at Stanford about incremenental improvement of data (in particular fiction) derived from the Google or Open Content Alliance projects. This sits outside the direct reponsibilities of Monk, but any 'good enough' public domain texts created in this fashion will clearly be useful to Monk.

Catherine observed that the use cases, story boards, and functional requirements developed in the Uses and Users Cell of the wiki haven't been studied as closely as they should. We agreed that a thorough review of these materials is critical to determining work flow and priorities for the next few months and that at the next SuperCell meeting we would be able to make decisions about priorities.

There was some discussion about how best to ensure that decisions taken in one cell would always be made in full awareness of consequences for other cell. There seemed agreement that keeping an eye on this was the responsibility of the super cell and that an additional layer of formal monitoring was not likely to do much good.