Public Classifiers

This is a directory of classifiers that have been published by users.

Public classifiers (38)

Sentiment

This classifier determines if a text is positive or negative. It can be used to conduct market research, brand surveys and see trends around campaigns. It's based on about 40000 Amazon product reviews (collected by Mark Dredze cs.jhu.edu/~mdredze/datasets/sentiment/) and mood training data (by mattiasostmar.net). You can read more in our blog:
34692941
total classifications

Text Language

Classifies the language of a text by looking on about 4000 commonly used words per language. It works best with clean texts but can also be used for HTML pages. For reliable results HTML pages need more text content (since HTML often contains English words and comments).
19807643
total classifications

Topics

Categories an English text into a topic (Arts, Business, Computers, Games, Health, Home, Recreation, Science, Society and Sports). Each of those topics has more specific child classifier (Art Topics, Business Topics etc). It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
3349663
total classifications

GenderAnalyzer_v5

This classifier tries to figure out if a text is written by a male or female. It has been trained on 11000 blogs (5500 blogs wrtten by females and 5500 by males). More text gives better results.
3203087
total classifications

Mood

The state of mind of the writer - upset or happy. On the extreme side there is angry, hateful writers and on the other extreme there is joyful and loving writers. The measured accuracy is 96% (using 10-fold cross validation). For reliable results we recommend that you use at least 200 words. Currently works for English texts.
By prfekt
3117454
total classifications

Ageanalyzer

This classifier tries to estimate to which age group a blog belongs. The training data is based upon about 7000 blogs collected randomly from Internet.
2188051
total classifications

Myers Briggs Judging Function

Determines the Thinking/Feeling dimension of the personality type according to Myers-Briggs personality model. The analysis is based on the writing style and should NOT be confused with the MBTI (c) which determines personality type based on self-assessment questionnaires. Currently works for English texts. Swedish version available on request.
By prfekt
1405330
total classifications

emo

By saifer
1329825
total classifications

classics

This classifier has been trained on 21 different classical authors. We have used about three books per author collected from the Gutenberg project. It only works for English texts. Try it out to see which poet your blog or text is most alike!
1135389
total classifications

Society Topics

Categories an English text into a society topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
1012682
total classifications

Tonality

Determines the tonality of a text - corporate (formal) or personal (informal). Helps distinguish between prosumer media and pro media for instance. Currently works for English texts.
By prfekt
744570
total classifications

News Classifier

This classifier categorizes news articles. It was trained with 1000 hand picked articles from major news sources per category. The text used was a combination of each article's title, description and scraped and cleaned text.
727473
total classifications

Myers Briggs Lifestyle

Determines the Judging/Perceiving dimension of the personality type according to Myers-Briggs personality model. The analysis is based on the writing style and should NOT be confused with the MBTI (c) which determines personality type based on self-assessment questionnaires. Currently works for English texts. Swedish version available on request.
By prfekt
510099
total classifications

Myers Briggs Attitude

Analyzes the Extraversion/Introversion dimension of the personality type according to Myers-Briggs personality model. The analysis is based on the writing style and should NOT be confused with the MBTI (c) which determines personality type based on self-assessment questionnaires. Currently works for English texts. Swedish version available on request.
By prfekt
506009
total classifications

quotes

By virus32
459853
total classifications

Home Topics

Categories an English text into a home topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
161708
total classifications

Art Topics

Categories an English text into an art topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
123359
total classifications

Game Topics

Categories an English text into a game topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
84902
total classifications

Health Topics

Categories an English text into a health topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
84745
total classifications

Computer Topics

Categories an English text into a computer topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
79112
total classifications

Business Topics

Categories an English text into an business topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
75585
total classifications

Recreation Topics

Categories an English text into a recreation topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
73786
total classifications

Sport Topics

Categories an English text into a sport topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
61029
total classifications

Science Topics

Categories an English text into a science topic. Use the parent classifier 'Topics' to find out if a text belongs in this category. It uses a subset of topics from the Open Directory Project at http://www.dmoz.org.
55230
total classifications

Myers Briggs Perceiving Function

Determines the Sensing/iNtuition dimension of the personality type according to Myers-Briggs personality model. The analysis is based on the writing style and should NOT be confused with the MBTI (c) which determines personality type based on self-assessment questionnaires. Currently works for English texts. Swedish version available on request.
By prfekt
28786
total classifications

Example News Categorizer

This news categorizer is only a simple example used in a tutorial. It has been trained on 20 texts per category (sports, entertainment and science) so don't expect too much (even though it seems to do incredibly well).
16250
total classifications

Punner

Classification between Semantic/pragmatic puns and phonological puns
By aangtce
16026
total classifications

myhumanfilter

there are three types of websites in the world.
274
total classifications

typeology

252
total classifications

tags

kvista's librarything classifier
By kvista
209
total classifications

CONTROLE

By virus32
198
total classifications

Apple

To tell the difference between Apple the Fruit and Apple Computer
168
total classifications

MySpamFilter

By khodrog
143
total classifications

whoareyou

129
total classifications

ZanrMuzikePoTekstu

Classifies music into Rock or Pop based on lyrics. A few hundies of lyrics were used to train classifier. Ovaj klasifikator ce razvrstati muziku po žanrovima na osnovu teksta. Uspešnost zagarantovana!
By fon
122
total classifications

gender

www.MyTwiShirt.com is your only source for twiShirts. A "twishirt" is a t-shirt with your Twitter username and photo on it (just in case you didn't guessed it already :). The process is really simple, you sign in with your Twitter account and get your twiShirt instantly. During the process we automatically detect your gender based on your tweets, and you're giving you 10% price deal for just one tweet. The shipping is done by Zazzle, world leader in print-on-demand services.
115
total classifications

Trond Bie

By kaihv87
30
total classifications

Tennis Equipment

30
total classifications