fast_psa: Probabilistic Latent Semantic Analysis
In svs: Tools for Semantic Vector Spaces

Description Usage Arguments Details Value References Examples

A fast procedure for computing probabilistic latent semantic analysis.

fast_psa(dat, k, symmetric = FALSE, transform = 1, tol = 1e-08)

fast_psi(dat, k, symmetric = FALSE, transform = 1, tol = 1e-08)

fast_plsa(dat, k, symmetric = FALSE, transform = 1, tol = 1e-08)

fast_plsi(dat, k, symmetric = FALSE, transform = 1, tol = 1e-08)

`dat`	Input data: can be a table or a data frame (but the data frame must have only two columns).
`k`	Numeric specification of the number of latent classes to compute.
`symmetric`	Logical indicating whether to compute the symmetric or the asymmetric solution.
`transform`	Numeric specification of the "tempering" transformation as explained in Hofmann (1999: 51-52).
`tol`	Numeric specification of the convergence criterion.

From version 1.1.0 of the svs package on, probabilistic latent semantic analysis is a special case of latent class analysis.

A list with components:

`prob0`	The probabilities of the latent classes.
`prob1`	The probabilities for the first set of levels (viz. the row levels of a frequency table). The rows of `prob1` sum to 1 if `symmetric = FALSE`, the columns sum to 1 if `symmetric = TRUE`.
`prob2`	The probabilities for the second set of levels (viz. the column levels of a frequency table). The columns of `prob2` sum to 1.

Hofmann, Th. (1999). Probabilistic latent semantic indexing. SIGIR'99: Proceedings of the 22nd annual international SIGIR conference on research and development in information retrieval, 50–57.

SndT_Fra <- read.table(system.file("extdata", "SndT_Fra.txt", package = "svs"),
   header = TRUE, sep = "\t", quote = "\"", encoding = "UTF-8",
   stringsAsFactors = FALSE)
psa.SndT_Fra <- fast_psa(SndT_Fra, k = 7)
psa.SndT_Fra

$prob0
[1] 0.22995445 0.22494661 0.21102856 0.13517149 0.10818295 0.06968235 0.02103360

$prob1
                           1             2             3          4
aanvang         0.000000e+00  1.147010e-98  3.835355e-04 0.00000000
begin          5.220146e-287  1.232633e-25  1.000000e+00 0.00000000
beginnen        2.746268e-02  7.956513e-01  4.212875e-02 0.01215805
eerst           9.376642e-01  4.137597e-02  2.095987e-02 0.00000000
gaan           6.712619e-285  1.235050e-01  0.000000e+00 0.00000000
komen           1.203333e-74  9.102385e-16 1.669604e-290 0.00000000
krijgen         1.488922e-01 3.751130e-211  0.000000e+00 0.00000000
ontstaan        6.500000e-01 1.973994e-160  1.672486e-01 0.00000000
openen          1.000000e+00 9.654997e-132  0.000000e+00 0.00000000
oprichten       0.000000e+00  1.076755e-71  0.000000e+00 0.00000000
opstarten      1.897119e-305  4.555378e-11 3.349381e-280 0.00000000
opzetten        5.029269e-01  2.074925e-16  0.000000e+00 0.00000000
start           0.000000e+00  3.622206e-95  8.619168e-01 0.00000000
starten         6.095142e-02  4.767526e-01  3.833153e-05 0.02000000
van start gaan 1.506560e-314  2.610802e-01 2.163344e-243 0.00000000
vanaf           0.000000e+00  0.000000e+00  0.000000e+00 1.00000000
worden          0.000000e+00  5.807865e-01  0.000000e+00 0.00000000
                           5             6             7
aanvang        5.420844e-128  4.160617e-02  9.580103e-01
begin           1.224847e-81 2.541888e-161  5.780879e-22
beginnen        1.082085e-01  1.223045e-02  2.160314e-03
eerst          3.055509e-227 4.091921e-230  1.266778e-28
gaan            7.800066e-01  9.648843e-02  0.000000e+00
komen           5.216735e-24  1.000000e+00 4.508937e-290
krijgen         1.564432e-01  6.946646e-01  0.000000e+00
ontstaan       3.553722e-205 1.099439e-282  1.827514e-01
openen          8.416589e-36 5.325830e-115  0.000000e+00
oprichten       1.000000e+00 1.015665e-128  5.662009e-35
opstarten       1.000000e+00 3.957306e-111 8.123207e-102
opzetten        4.970731e-01 5.684401e-140  0.000000e+00
start           1.380832e-01 1.058162e-136  3.635909e-18
starten         4.002108e-01  1.243751e-02  2.960936e-02
van start gaan  6.489076e-01  9.001223e-02 5.148050e-115
vanaf           0.000000e+00  0.000000e+00  0.000000e+00
worden          5.792672e-25  4.192135e-01  1.755465e-79

$prob2
                             1            2             3             4
<U+00E0> partir de 4.663579e-290 1.480936e-09 8.039300e-255  1.000000e+00
commencer         2.727142e-02 6.274856e-01  1.021747e-02 7.548992e-289
d'abord           5.264036e-01 1.043558e-06 3.868479e-215 2.185659e-304
d<U+00E9>but        4.484207e-41 7.827707e-09  9.746842e-01 4.226717e-290
d<U+00E9>butant    9.272764e-260 2.989576e-02 4.736897e-258 5.562744e-270
d<U+00E9>buter     4.803383e-295 1.314945e-01  1.149707e-02 3.956595e-291
d<U+00E9>marrer    3.439553e-245 5.356274e-04 6.934479e-264 2.950312e-268
entamer          2.765778e-201 1.075287e-01 1.486189e-230 2.248875e-283
entreprendre      0.000000e+00 9.048784e-03 6.051877e-232 5.164837e-307
entrer           1.082745e-290 1.609449e-08 6.965526e-245 3.967685e-294
lancer            3.084002e-03 8.701817e-09  3.601260e-03 2.125848e-295
lancer, se       2.625972e-176 4.145958e-02 9.256895e-253 8.104435e-305
mettre, se       2.667954e-322 2.495021e-02 2.833912e-238 1.664156e-314
ouvrir            4.432410e-01 7.833559e-04 2.429861e-236 7.487267e-276
partir           6.072495e-316 2.010765e-10 3.455134e-138 1.722331e-304
prendre cours     0.000000e+00 1.186889e-02 5.632793e-221 2.915840e-303
recommencer       0.000000e+00 1.494788e-02 2.752016e-255 3.952525e-323
                             5             6             7
<U+00E0> partir de  1.666026e-62 1.685130e-246 9.630925e-212
commencer         3.081703e-02  4.458554e-02 2.568120e-165
d'abord           5.005686e-93 1.582930e-254 1.321238e-266
d<U+00E9>but        5.935367e-62 7.323496e-234  7.080167e-01
d<U+00E9>butant     1.332966e-27 1.052884e-232 2.656229e-174
d<U+00E9>buter      9.288251e-02  3.281865e-02  3.406582e-02
d<U+00E9>marrer     2.257621e-01  2.415537e-02 1.597214e-168
entamer           6.213263e-02 6.293626e-126  2.249888e-01
entreprendre      2.469864e-02 9.177636e-160 3.285326e-303
entrer            7.234871e-23  7.527674e-01  1.370128e-15
lancer            2.633997e-01  6.217839e-02 3.614493e-178
lancer, se        2.568536e-02  4.289907e-64 6.964590e-241
mettre, se        1.129988e-01  7.215336e-02 1.559790e-310
ouvrir            1.266637e-32  1.134124e-02  2.499482e-68
partir            1.616232e-01 1.041371e-153 7.797240e-229
prendre cours     3.812971e-98 9.844292e-231  3.292864e-02
recommencer      2.922057e-114  2.437341e-60 7.539936e-320