Query Results

Number of results 10000
Lowest scoring result 129.08723
Abstracts of results results.html
Complete abstracts of the articles in Medline predicted to be relevant (limited number per page).
Abstracts of all results ZIP file or all_results.html
All result abstracts on a single page. (can take a long time to load - downloading the zip file is recommended)
Abstracts of input examples inputs.html
Complete abstracts of the Medline records given as relevant training examples. They have been ranked by classifier scores.
PubMed IDs of results results.txt
PubMed IDs and scores of classifier predictions, ranked by decreasing score.
PubMed IDs of inputs inputs.txt
PubMed IDs and scores of the relevant training examples.
Started at 2009/09/29 13:16:00 GMT
Time at which query was started
Finished at 2009/09/29 13:23:26 GMT
Time at which this file was written.
Feature score method scores_laplace_split
Name of the method used to calculate feature scores. Docstring for the method: For feature probabilities we use a Laplace prior, of 1 success and 1 failure in total, split between the classes according to size. This avoids problems with class skew.
Min Information Gain 2e-05
We exclude features with less than this value of Information Gain.
Base score -63.5576964411
The log likelihood ratio of an empty article (one in which every feature failed to occur).
Prior score -8.16379067647
The log of the prior probability ratio for an article being relevant versus irrelevant (added to log likelihood ratio to obtain the final score). Equals the logit of the estimated prevalence of relevant articles in Medline (which may be estimated from the input size or specified separately).
Limit 10000
The maximum number of results to include.
Threshold -30.0
Default Naive Bayes classification threshold is zero. This threshold is the minimum log probability ratio for predicting an article to be relevant.

Feature Statistics

Quantity Relevant Docs Irrelevant Docs
Number of documents 4848 17027128
Number of selected, occurring features 40502 56005
Total occurrences of selected features 490927 764476920
Selected features per Medline record 101.264 44.898
Of the considered feature types, 56022 features are selected out of 3703762 occurring at least once in training data. The aggressivity of selection is 66.113. The complete database lists 3703762 potential features.

Features with high TF.IDF

Features with TF.IDF above 0.2 or 0.3 could make good keywords. TF.IDF is term frequency times inverse document frequency, where we treat the set of input citations as a single document

TF-IDF Type Term Term ID Score Pos Neg
0.05 w Melanoma 3153 7.56 4093 48143
0.04 mesh Melanoma 16413 6.41 3067 48187
0.02 w Genes 1395 3.95 2641 382374
0.02 qual genetics 799 3.91 4109 1714509
0.02 w Gene 1420 3.39 2780 738983
0.02 w Expression 2275 3.08 2517 807958
0.02 w Tumor 161 3.24 2077 483905
0.01 w Melanomas 17486 6.02 907 9503
0.01 w Cell 384 2.60 2699 1457268
0.01 mesh Tumor Cells, Cultured 182 3.83 1362 143073
0.01 w Cells 213 2.48 2812 1758532
0.01 mesh Skin Neoplasms 5098 4.31 1151 71011
0.01 w Human 442 2.45 2328 1256660
0.01 w Malignant 6944 3.35 1234 202117
0.01 w Lines 2830 3.32 1193 198898
0.01 w Cancer 167 2.49 1507 611007
0.01 w Tumors 160 2.93 1192 290132
0.01 mesh Gene Expression Regulation, Neoplastic 30142 4.40 719 36337
0.01 mesh Melanoma, Experimental 42021 6.06 541 4969
0.01 w Metastatic 2064 3.62 813 91652