Performance: Record of an Activity Trace
In mpia: Meaningful Purposive Interaction Analysis

Description Usage Arguments Details Value Warning Author(s) References See Also Examples

The Performance holds a digital record representing a particular human action: it stores a representation of meaning and an human-readable label for the purpose of the action. The purpose thereby is a character string, whereas the meaning vector is constructed by mapping the descriptive source text to its social semantic vector space representation.

Several class methods (and generics) support in inspecting the result of such mapping: terms, for example, returns all terms activated above threshold in the vector space through the fold in of the source text.

Each Performance object is bound to a particular Domain.

1 2	Performance( text=NULL, purpose=NULL, domain=NULL, name=NULL, weighting=NULL, logging=TRUE )

`text`	Character string containg the text written or read (or the descriptive text for any other sorts of actions).
`purpose`	Character string describing the purpose of the action, e.g. "course 101 exam".
`domain`	Reference pointing to a `Domain` object.
`name`	Character string labelling the performance (e.g. "Essay for business management exam").
`weighting`	A weighting function to be executed on the raw query vector (before projection into the Eigenspace).
`logging`	Logical flag indicating whether the source text shall be internally stored in field `sourcetext`.

Any Performance is associated to a particular Domain. It is constructed from a given source text (which may or may not be internally stored, depending on the setting for logging) by mapping the text to its social semantic vector space representation. This projection uses the fold-in facilities provided by the Domain object. For efficiency reasons, the projection appends the meaning vector in the domain object in the traces matrix and returns its index position to be held by the performance record.

The constructor 'initialize' returns the reference pointing to the object.

Using weighting functions is complicated!

Fridolin Wild <wild@brookes.ac.uk>

Fridolin Wild (2016): Learning Analytics in R with SNA, LSA, and MPIA, Springer:Berlin.

More detailed information: Performance-class, terms, performances, overlap, plot, Domain-class, Domain

d = Domain(name="test")
evidence = TermDocumentMatrix( Corpus( VectorSource( c("abc abc def",
    "def ghi", "ghi abc, lmno", "lmno abc wxyz") ) ) )
d$corpus(evidence)
d$spacify()

p = Performance(text="abc def ghi", domain=d, purpose="demo", name="demo")
terms(p)