Page 133 - JOURNAL OF LIBRARY SCIENCE IN CHINA 2018 Vol. 42
P. 133
132 Journal of Library Science in China, Vol. 8, 2016
number of the relevant documents. The precision and recall of the two-round retrieval experiments
for the library OPAC system are shown in Figure 5 and Figure 6 respectively.
3
3
᛫
ѹ η η ሙ ሙ ஔ η ৱ ॷ ॷ ॎ ៶ ᛫ ኮ х ҷ ڎ ᇫ ࠫ ܱ η ᝠ ေ ሙ ඟ ҷ ၷ
ழ ৌ ৌ ߦ ߦ ৌ ਖ ေ ေ ٨ र Ἧ ေ ኖ ү ࠒ ͘ ܱ ̔ ৌ ካ ߦ ீ ү ေ
ኮ ೝ ߦ ᆑ ੇ ေ Ϥ ೝ ॷ ܫ ߦ ߦ ҧ ᛡ ˟ Т ܫ ኮ வ ҧ ߦ Ӽ ߦ
ေ ጊ ቃ ౧ ష क ጊ ᣤ ေ ေ ˧ ኖ ጇ ေ ေ ข ߦ ѳ ၷ
ߦ ᭩
ἰ Т ֑
Note：R1 is the recall without the use of the terminology service, R2 is the recall with the use of the terminology service.
ฌ3 ˞ళΎၹషឦҬᄊಊЛဋὊ3 ˞ΎၹషឦҬᄊಊЛဋ
Figure 5. The recalls of the library OPAC system with or without the use of the “GetSynonym” terminology service.
1
1
᛫
ѹ η η ሙ ሙ ஔ η ৱ ॷ ॷ ॎ ៶ ᛫ ኮ х ҷ ڎ ᇫ ࠫ ܱ η ᝠ ေ ሙ ඟ ҷ ၷ
ழ ৌ ৌ ߦ ߦ ৌ ਖ ေ ေ ٨ र Ἧ ေ ኖ ү ࠒ ͘ ܱ ̔ ৌ ካ ߦ ீ ү ေ
ኮ ೝ ߦ ᆑ ੇ ေ Ϥ ೝ ॷ ܫ ߦ ߦ ҧ ᛡ ˟ Т ܫ ኮ வ ҧ ߦ Ӽ ߦ
ေ ጊ ቃ ౧ ష क ጊ ᣤ ေ ေ ˧ ኖ ጇ ေ ေ ข ߦ ѳ ၷ
ߦ ᭩
ἰ Т ֑
Note: P1 is the precision without the use of the terminology service, P2 is the precision with the use of the terminology service.
ฌ1 ˞ళΎၹషឦҬᄊಊюဋὊ1 ˞ΎၹషឦҬᄊಊюဋ
Figure 6. The precisions of the library OPAC system with or without the use of the “Getsynonym” terminology service.
As shown in Figure 5, the recall of all the experiments increases after calling the terminology
service. The average recall of the first-round experiments is 50.3%, whereas the average recall
of the second-round is 77% with an increase of 26.7%. This means that expanding query terms
through the “GetSynonyms” terminology service can significantly improve the completeness
of retrieval results. As shown in Figure 6, the average precision of the first-round experiments
is 64.4%, whereas the average precision of the second-round is 59.8% with a decrease of 4.6%.
In comparison to the consistent improvement of all the experiments’ recall values, the precision
values vary differently. The precision of only a third of the experiments increases after calling the
