iSA: The iSA algorithm

In blogsvoices/iSAX: This Package Implements iSA Algorithm *

The iSA algorithm

Description

This function implements the iSA - integrated Sentiment Analysis algorihtm

Usage

iSA(Strain, Stest, Dtrain, nboot = 1000, predict = FALSE,
    ret.boot = FALSE, seqlen = 5, sparse = FALSE,
    verbose = TRUE, tolerance=Inf)

Arguments

Strain	a vector of stem strings belonging to the training set
Stest	a vector of stem strings belonging to the test set
Dtrain	a vector of codings belonging to the training set
nboot	number of bootstrap replications for standard error estimation
predict	non in use
ret.boot	return all bootstrap estimates?
seqlen	length of substrings used in iSA
sparse	use sparse matrix to store data?
verbose	show all steps?
tolerance	threshold for numerical determinant. Defaut = 0 means ignore it.

Details

Prediction is implemented but only beta. Use it at your own risk.

Value

est	estimated parameters
tab	table of results
best	bootstrap estimated parameters
btab	table of results for bootstrap estimates
pred	prediction for each entry
tim	execution time

Author(s)

Stefano M. Iacus

References

Iacus, S.M., Curini, L., Ceron, A. (2015) iSA (U.S. provisional patent application No. 62/215264) Ceron, A., Curini, L., Iacus, S.M. (2016) iSA: A fast, scalable and accurate algorithm for sentiment analysis of social media content, Information Sciences, V. 367-368, p. 105-124.

See Also

prep.data

blogsvoices/iSAX documentation built on Oct. 11, 2022, 2:38 p.m.