hjp
/
bayes
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Avoid overlapping tokens 

For each used token, record the first, second and last third and exclude
all tokens which include those.
This commit is contained in:
Peter J. Holzer 2019-08-17 11:12:34 +02:00
parent f3817c4355
commit 631f97abe5
1 changed files with 14 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
judge_message
View File

@ -10,6 +10,8 @@ import psycopg2.extras as ppe
components = {}
EVIDENCE_MAX = 20
def add_component(t, c):
if t not in components:
components[t] = []
@ -55,24 +57,31 @@ def extract_features():
csr.execute(q, (t, length, *current))
for i, r in enumerate(csr):
prev.add(r["feature"])
if i < 15 and r["interesting"]:
if i < EVIDENCE_MAX and r["interesting"]:
evidence.append(r)
length += 1
evidence = sorted(evidence, key=lambda x: -x["length"])
evidence = sorted(evidence, key=lambda x: -x["interesting"])
interesting_evidence = []
seen = []
for e in evidence:
for i in interesting_evidence:
if e["type"] == i["type"] and e["feature"] in i["feature"]:
for s in seen:
if e["type"] == s["type"] and (e["feature"] in s["feature"] or s["feature"] in e["feature"]):
break
else:
interesting_evidence.append(e)
if len(interesting_evidence) >= 15:
seen.append(e)
ln = len(e["feature"])
if ln >= 6:
seen.append({"type": e["type"], "feature": e["feature"][0 : ln // 3]})
seen.append({"type": e["type"], "feature": e["feature"][ln // 3 : ln * 2 // 3]})
seen.append({"type": e["type"], "feature": e["feature"][ln * 2 // 3 : ln]})
if len(interesting_evidence) >= EVIDENCE_MAX:
break
p1 = 1
p2 = 1
for i in interesting_evidence:
print("#", i["spam_prob"], i["type"], i["feature"])
print("#", i["spam_prob"], i["type"], i["length"], i["feature"], sep="\t")
p1 *= i["spam_prob"]
p2 *= 1 - i["spam_prob"]
p = p1 / (p1 + p2)