David M. Blei (2012): Probabilistic Topic Models. Communications
of the ACM, Vol. 55, No. 4, p. 78.
| 0 | katholisch religion könig katholik kaiser protestantisch handeln niederland provinz protestantismus |
| 6 | erziehung erzieher tugend mensch ideell zögling gesellschaft abhandlung historisch inner |
| 7 | grenze nationalität heutig sprachgrenze südlich süddeutsch gebiet inner muttersprache einheitsstaat |
| 14 | bahn bahnlinie konzession bahnbau projekt geplant annullierung britisch anwenden ausdrücklich |
| 27 | virtuose pianoforte talent vater virtuos knabe welt klavierlehrer portrait instrument |
| Name | Developer | Language | Link | ||
|---|---|---|---|---|---|
| MALLET | machine learning for language toolkit | ![]() |
Andrew McCallum et al. | Java | http://mallet.cs.umass.edu/topics.php |
| Gensim | topic modeling for humans | ![]() |
Radim Řehůřek | Python | https://radimrehurek.com/gensim |
| tmw | topic modeling workflow | ![]() |
Christof Schöch | Python | https://github.com/cligs/tmw |
| dfr-browser | a simple topic-model browser | ![]() |
Andrew Goldstone | JavaScript | http://agoldst.github.io/dfr-browser/ |
|
|
|
|
"The words returned from the model paint both setting and subject; the result is a useful and quantifiable representation of a very particular theme [...] These are macro trends we are exploring, and they provide a generalized view of the whole"
| 4 | de la el en green verde con los mi se |
| 6 | blue red white bird color green yellow black wings birds |
| 9 | thy thou thee art thine st doth heaven hast hath |
| 32 | night light moon stars day dark sun sleep sky wind |
| 54 | tree green summer flowers grass trees flower spring leaves sun |
| 58 | gertrude guitar inside blue stein beginning sieve cloud type end |
"My research confirms, to a degree, Ted Underwood’s suspicion that topics in literary studies are better understood as a representation of “discourse” (language as it is used and as it participates in recognized social forms) rather than a thematic string of coherent terms."
|
|
|
|
"The latter types of topics [...] show that taking a method such as Topic Modeling, developed initially for collections of non-fictional prose such as scholarly journal articles or newspapers, and adapting it to the domain of literary texts, actually changes the meaning of the word 'topic'"
Folien: https://hennyu.github.io/dgavl_17
CLiGS-Gruppe: http://cligs.hypotheses.de/