def __init__(self):
self.dataset = FileIO.read_json_file("sematch/benchmark-data/data.txt")
self.results = FileIO.read_list_file("sematch/benchmark-data/result.txt")
self.queries = [(d['query'],d['entity']) for d in self.dataset]
self.relevants = [d['result'] for d in self.dataset]
self.synsetExpansion = SynsetExpansion()
self.engine = Engine()
self.sims = ['wup', 'lch', 'res','jcn', 'lin']
self.thresholds = [0.9,1]
self.gpcs = ['gpc1', 'gpc2', 'gpc3', 'gpc4', 'gpc5', 'gpc6']
from flask import Flask, jsonify, json, request, render_template as template
from QueryEngine import Engine
import os
DEBUG = True
SECRET_KEY = 'Secret_development_key'
app = Flask(__name__)
app.config.from_object(__name__)
engine = Engine()
@app.route('/api/queries')
def queries():
query = request.args.get('query')
return json.dumps(engine.type_entity_query_construction(query))
@app.route('/api/types')
def types():
query = request.args.get('query')
return json.dumps(engine.types(query))
@app.route('/api/type_search')
def type_search():
type = request.args.get('type')
sim = request.args.get('sim')
th = request.args.get('th')
result = engine.search_types(type, sim, float(th))
class Experiment:
def __init__(self):
self.dataset = FileIO.read_json_file("sematch/benchmark-data/data.txt")
self.results = FileIO.read_list_file("sematch/benchmark-data/result.txt")
self.queries = [(d['query'],d['entity']) for d in self.dataset]
self.relevants = [d['result'] for d in self.dataset]
self.synsetExpansion = SynsetExpansion()
self.engine = Engine()
self.sims = ['wup', 'lch', 'res','jcn', 'lin']
self.thresholds = [0.9,1]
self.gpcs = ['gpc1', 'gpc2', 'gpc3', 'gpc4', 'gpc5', 'gpc6']
@staticmethod
def measure(relevant, retrieved):
a = len(relevant)
b = len(retrieved)
ab = 0
for re in retrieved:
if re in relevant:
ab += 1
recall = float(ab) / float(a)
if b == 0:
precision = 0
else:
precision = float(ab) / float(b)
if precision + recall == 0:
f = 0
else:
f = 2 * (precision * recall) / (precision + recall)
return recall, precision , f
def query_info(self, id):
tQ, eURI = self.queries[id]
relevant_data = self.relevants[id]
print id, '\tquery: ', ' '.join(self.dataset[id]['terms'])
print tQ, eURI
print 'N(relevant)=', len(relevant_data)
def print_query(self, id, gpc, sim, th):
tQ, eURI = self.queries[id]
tURIs = self.typeExpansion.expandType(tQ, sim, th)
print "Number(types)=", len(tURIs)
query = self.engine.query(gpc, tURIs, eURI)
#result = self.engine.sparql.execute(query)
result = self.engine.sparql.request_execution(query)
print result
def experiment(self, id, gpcs, sim, th):
tQ, eURI = self.queries[id]
relevant_data = self.relevants[id]
tURIs = self.typeExpansion.expandType(tQ, sim, th)
result = self.engine.run(gpcs,tURIs, eURI)
#print sim, th, gpcs
#print "N(type uris)=",len(tURIs)
#print 'N(returned)=', len(result)
#print 'recall is %f, precison is %f, f measure is %f' % Experiment.measure(relevant_data,result)
print Experiment.measure(relevant_data,result)
def analysis(self):
data = self.results
data = [d.lstrip('(') for d in data]
data = [d.rstrip(')') for d in data]
data = [d.split(',') for d in data]
data = [map(float, d) for d in data]
data = [data[i:i+20] for i in range(0,len(data),20)]
sum_data = data[0]
for d in data[1:]:
for i in range(20):
sum_data[i][0] += d[i][0]
sum_data[i][1] += d[i][1]
sum_data[i][2] += d[i][2]
for i in range(20):
print i+1, '***************'
print sum_data[i][0] / 22.0
print sum_data[i][1] / 22.0
print '***************'