README.md

WPP2 - Gruppe 7

Install

pip install -r requirements.txt

Usage

The program is an interactive shell. Start with

python main.py

Commands

search '<Query>'

Info

To use Proximity Queries use a forward slash '/' and not '\' because the "click" package parses it wrong

Benchmarks

• Setup index: 3.41s

• Search search '(blood OR pressure) AND cardiovascular': 0.001s

• Search search '(blood OR presure) AND cardiovascular' (with spell check): 0.011s

New Benchmarks

• Setup (Load word2vec, index, load vectors): 1min 43sec

Evaluation

PLAIN-121

Method	Precision@5	Precision@10	Precision@20	Precision@50	Recall@5	Recall@10	Recall@20	Recall@50	F1@5	F1@10	F1@20	F1@50	R-Precision
td-idf	0.8	0.4	0.2	0.12	0.05714285714285714	0.05714285714285714	0.05714285714285714	0.08571428571428572	0.1066	0.0999	0.0888	0.0999	0.08571428571428572
word2vec	0.6	0.6	0.3	0.16	0.04285714285714286	0.08571428571428572	0.08571428571428572	0.11428571428571428	0.08	0.15	0.1333	0.1333	0.11428571428571428

PLAIN-1021

Method	Precision@5	Precision@10	Precision@20	Precision@50	Recall@5	Recall@10	Recall@20	Recall@50	F1@5	F1@10	F1@20	F1@50	R-Precision
td-idf	0.8	0.9	0.9	0.82	0.006779661016949152	0.015254237288135594	0.030508474576271188	0.06949152542372881	0.013445378151260505	0.0300	0.05901639344262295	0.128125	0.30338983050847457
word2vec	1.0	1.0	0.9	0.54	0.00847457627118644	0.01694915254237288	0.030508474576271188	0.04576271186440678	0.01680672268907563	0.0333	0.05901639344262295	0.08437499999999999	0.2559322033898305

PLAIN-15

Method	Precision@5	Precision@10	Precision@20	Precision@50	Recall@5	Recall@10	Recall@20	Recall@50	F1@5	F1@10	F1@20	F1@50	R-Precision
td-idf	0.2	0.2	0.2	0.1	0.02564102564102564	0.05128205128205128	0.10256410256410256	0.1282051282051282	0.0454	0.0816326530612245	0.13559322033898302	0.11235955056179775	0.1282051282051282
word2vec	0.0	0.2	0.15	0.08	0.0	0.05128205128205128	0.07692307692307693	0.10256410256410256	0.0	0.0816326530612245	0.1016949152542373	0.08988764044943821	0.10256410256410256

PLAIN-145

Method	Precision@5	Precision@10	Precision@20	Precision@50	Recall@5	Recall@10	Recall@20	Recall@50	F1@5	F1@10	F1@20	F1@50	R-Precision
td-idf	0.6	0.3	0.2	0.22	0.08571428571428572	0.08571428571428572	0.11428571428571428	0.3142857142857143	0.15	0.1333	0.1454	0.25882352941176473	0.22857142857142856
word2vec	0.0	0.0	0.05	0.08	0.0	0.0	0.02857142857142857	0.11428571428571428	0.0	0.0	0.0363	0.09411764705882354	0.08571428571428572

PLAIN-1336

Method	Precision@5	Precision@10	Precision@20	Precision@50	Recall@5	Recall@10	Recall@20	Recall@50	F1@5	F1@10	F1@20	F1@50	R-Precision
td-idf	0.2	0.2	0.1	0.04	0.125	0.25	0.25	0.25	0.15384615384615385	0.22222222222222224	0.14285714285714288	0.06896551724137932	0.125
word2vec	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0

MAP-Score

td-idf: 0.17417613459986342

word2vec: 0.11169926119078662