Vector-Space-Model-Search-Engine/eval.py at main · Monso0n/Vector-Space-Model-Search-Engine · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
from search2 import SearchEngine

a = SearchEngine() #instantiate search engine

queryFile = open("cacm/query.text", 'r') #import files for reading
relFile = open("cacm/qrels.text", 'r')

mode = None

queries = {}

q = ""
num = None

for i in queryFile: # get queries, put them in dict
    if i.startswith("."):
        mode = i.split()[0]

        if i.startswith(".I"):
            queries[num] = q
            q = ""

            num = int(i.split()[1])

    elif mode == ".W":
        q += i + " "

queries.pop(None)
print("\n")
#print(queries)

relevant = {}

num = 0
for i in relFile: #get relevant files, put them in dict
    num = int(i.split()[0])
    docId = int(i.split()[1])

    #print(f"{num}  {docId}")

    if num not in relevant:
        relevant[num] = [docId]
    elif num in relevant:
        list = relevant[num]
        list.append(docId)

        relevant[num] = list


def MAP(ret, rel): #returns average precision of top 10

    precisionSum = 0
    relevantCount = 0
    position = 0

    for i in ret:
        position+=1

        if i in rel:
            relevantCount+=1

            prec = relevantCount/position
            print(f"'{i}' is at index {position}, precision for retrieved doc is: {prec} = {relevantCount} / {position}")
            precisionSum+=prec


    AP = precisionSum / len(rel)
    print(f"\nThe Average Precision is: {AP} = {precisionSum} / {len(rel)}")
    return AP


def rPrec(ret, rel):

    count = 0
    size = len(rel)


    if len(rel) > len(ret):
        print("OVER HERE")
        for i in rel:
            if i in ret:
                count+=1
    else:
        for i in range(size):
            if ret[i] in rel:
                count+=1


    rprecision = count/size
    print(f"{count} relevant results in top {size} retrieved")
    print("\nThe R-Precision is: {} = {} / {}\n".format(rprecision,count,size))
    return rprecision


rprecisionList = []
APList = []

for k, v in relevant.items():

    query = queries[k]
    print("--------------------------------------------------------------------------------")
    print(f"Query Number: {k}")
    rel = relevant[k]

    returned = a.search(query, False)

    retrieved = []
    for i in returned:
        retrieved.append(i[0])


    print(f"query(shortened): {query[:150]}")
    print("relevant docIDs: {}".format(rel))
    print("retrieved docIDs: {}\n".format(retrieved))

    prec = rPrec(retrieved, rel)
    AP = MAP(retrieved, rel)

    APList.append(AP)
    rprecisionList.append(prec)
    print("\n")


print("\n----------------------------------------------")

print("All queries processed\nAverage Results")
print(f"Average R-Precision: {sum(rprecisionList) / len(rprecisionList)}")
print(f"Mean Average Precision: {sum(APList) / len(APList)}")