Query Results

Number of results 0
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 2009/05/14 07:36:51 GMT
Time at which query was started
Finished at 2009/05/14 07:39:56 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 -194.794763417
The log likelihood ratio of an empty article (one in which every feature failed to occur).
Prior score -15.2643085979
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 1000
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 3 17031973
Number of selected, occurring features 277 29774
Total occurrences of selected features 326 812147890
Selected features per Medline record 108.667 47.684
Of the considered feature types, 29774 features are selected out of 3703762 occurring at least once in training data. The aggressivity of selection is 124.396. 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
0.09 a tp flores 1812890 14.78 2 11
0.07 a ca orengo 50989 12.67 2 105
0.05 mesh Software 23437 21.47 3 51614
0.05 issn 0961-8368 136829 8.93 2 4523
0.05 w MULTAL 3455448 14.57 1 2
0.05 w CATHEDRAL 1826857 14.57 1 2
0.05 w Double-dynamic 1249772 14.17 1 4
0.05 mesh Models, Molecular 1777 20.65 3 116419
0.05 a t dallman 1826855 13.88 1 6
0.04 w Folds 66737 8.00 2 11454
0.04 a fm pearl 1826856 13.32 1 12
0.04 w Embellishments 1007430 12.70 1 24
0.04 w Alignment 6361 7.28 2 23493
0.04 mesh Protein Folding 8397 7.21 2 25031
0.04 a dr westhead 1777525 12.07 1 47
0.04 w Structures 2516 19.91 3 242245
0.04 w Residue-based 407191 11.93 1 54
0.04 w SSAP 752823 11.90 1 56
0.04 w TOPS 421213 11.77 1 64
0.04 w Fold 392 6.88 2 34861