def get_es_results(abstracts, top_n):
es_results = []
config = Config()
search = Search()
for abstract in abstracts:
print(search.search_by_abstract(abstract, top_n, config))
es_results.append(search.search_by_abstract(abstract, top_n, config))
return es_results
def search_tpoic(topic, n=RETURN_SIZE):
s = Search(topic)
records = Record.records(topic, n)
if len(records) != 0:
results = [Result(record.doc_id, record.score) for record in records]
else:
results = s.search()
records = [
Record(pn=n,
topic_id=topic,
score=result.score,
doc_id=result.doc_id) for result in results
]
Record.save_all(records)
p = s.p(results, n)
res = format_results(topic, results, n, p)
return render_template('search_result.html', n=n, p=p, results=res)
def get_es_results(abstracts, top_n):
start_time = time.time()
es_results = []
config = Config()
search = Search()
for abstract in abstracts:
try:
result = search.search_by_abstract(abstract, top_n, config)
print(result)
print('搜索结果中包含 ' + str(len(result)) + ' 条数据')
es_results.append(result)
except (Exception) as e:
print('ES检索出现异常： Exception:', str(e))
end_time = time.time()
time_used = datetime.timedelta(seconds=int(round(end_time - start_time)))
print('检索耗时:' + str(time_used))
return es_results
def prepare():
"""
Elasticsearch config preparing, global arguments preparing, vocabulary preparing and model restore preparing.
:return:
"""
ir_config = Config()
search = Search()
args = path_arg4test()
# load vocab
print('Loading vocab...')
with open(os.path.join(args.vocab_dir, 'vocab.data'), 'rb') as fin:
vocab = pickle.load(fin)
print('Restoring the model...')
rc_model = RCModel(vocab, args)
rc_model.restore(model_dir=args.model_dir, model_prefix=args.algo)
return vocab, search, ir_config, args, rc_model
def search_by_user(disease, gene, n, type_):
s = Search()
results = s.search_by_user(disease, gene, type_)
#p = s.p(results, n)
res = format_results(None, results, n, p=None)
return render_template("search_result.html", n=n, p=False, results=res)
# from infer import prepare, inference, infer_prob
from infer_tune import prepare, inference, infer_prob
def chinese_tokenizer(documents):
"""
中文文本转换为词序列(restore时还需要用到，必须包含)
:param documents:
:return:
"""
for document in documents:
yield list(jieba.cut(document))
config = Config()
search = Search()
vocab_processor, model, session = prepare()
with open('../data/primary_question_dict.json') as primary_dict_f:
primary_question_dict = json.loads(primary_dict_f.readline())
class MatchHandler(tornado.web.RequestHandler):
def data_received(self, chunk):
pass
def get(self, *args, **kwargs):
self.render('find_answer.html')
def post(self, *args, **kwargs):
self.use_write()
from ir.search import Search s = Search(1) res = s.search() print(s.p(res, 10))