Identification and Analysis of Multi-tasking Product Information Search Sessions with Query Logs

Xiang Zhou; Pengyi Zhang; Jun Wang

Open Access

Identification and Analysis of Multi-tasking Product Information Search Sessions with Query Logs

Xiang Zhou

Pengyi Zhang

and

Jun Wang

| Sep 01, 2017

Journal of Data and Information Science

Volume 1 (2016): Issue 3 (August 2016)

About this article

Cite

Published Online: Sep 01, 2017

Page range: 79 - 94

Received: Mar 16, 2016

Accepted: Jun 06, 2016

DOI: https://doi.org/10.20309/jdis.201621

Keywords
Product search, Shopping task identification, Shopping task analysis, Multitasking session

© 2016 Xiang Zhou, Pengyi Zhang, Jun Wang

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Distribution of users by number of sessions.

Thresholds and F-scores of the clustering approaches.

Session with related search tasks.

SID	Time	Query terms	Query terms (translation)
1985	2013/5/20 20:21:34	休闲衬衫男短袖	Casual shirt male short sleeve
1985	2013/5/20 20:22:10	休闲衬衫男绿	Casual shirt male green
1985	2013/5/20 20:23:43	短裤男	Shorts male

Distribution of the sessions according to the number of task per session.

Number of task in a session	Freq.	Percent (%)	Cumulative percent (%)
1	2140	61.4	61.4
2	748	21.5	82.9
3	292	8.4	91.3
4	132	3.8	95.1
5	73	2.1	97.2
6	37	1.1	98.2
7	23	0.7	98.9
8	14	0.4	99.3
9	11	0.3	99.6
10	5	0.1	99.8
11 and more	8	0.2	100

Three related search tasks.

SID	Time	Query terms	Query terms (translation)
879	2013-05-04 12:06:25	增高鞋真皮休闲	Hidden heel shoes leather leisure
879	2013-05-04 12:20:25	夏季潮男洞洞鞋牛皮	Summer male leather crocs
879	2013-05-04 13:05:58	万斯低帮豹纹	Vance leopard print low-cut

Average session duration.

Session type	Average session duration
One task	36 minutes 9 seconds
Two tasks	54 minutes 19 seconds
Three or more tasks	1 hour 22 minutes 22 seconds

Average query length.

Session type	Average query length in characters
One task	7.56
Two tasks	7.28
Three or more tasks	7.32

Sample of human task identifications.

	Sid	Query Terms (original)	Query Terms (translation)	Coder
	Sid	Query Terms (original)	Query Terms (translation)
				#1	#2
1	1973	丰胸仪	Breast augmentation instrument	1	1
2	1973	优格格丰乳仪	Yougege breast augmentation instrument	1	1
3	1975	北京茶月饼	Beijing tea mooncake	2	2
4	1975	金凤呈祥	Jinfengchengxiang	3	3
5	1975	金凤呈祥 200	Jinfengchengxiang 200	3	3
6	1976	美优食品	Meiyou food	4	4
7	1977	XQB38-83皮带	XQB38-83 belt	5	5
8	1978	味多美卡	Meiduomei gift card	6	3
9	1978	Laver丰胸精油	Laver breast augmentation oil	7	6
10	1978	AOC 拉莫圣日尔曼干红葡萄酒 750ml	AOC Saint Germain Rameau claret 750ml	8	7
11	1978	AOC银奖圣玛杰庄园干红葡萄酒 750ml	AOC silver award Domaine Saint Majan claret 750ml	8	7
12	1978	圣玛杰庄园干红葡萄酒 750ml	Domaine Saint Majan claret 750ml	8	7
13	1978	红绳	Red rope	9	8
14	1978	红绳批发	Red rope wholesale	9	8
15	1978	项链挂绳编织	Necklace rope woven	9	8

Basic characteristics of sessions and tasks.

Item	Basic characteristics
Average number of queries per session	2.57
Highest number of queries in a session	21
Average number of tasks per session	1.78
Highest number of tasks in a session	41
Average number of queries per task	1.45
Highest number of queries in a task	15

Sample query records.

User ID	Sid	Query terms	Query terms (translation)
1028433974716967148	1973	丰胸仪	Breast augmentation instrument
1028433974716967148	1973	优格格丰乳仪	Yougege breast augmentation instrument
1028433974716967148	1975	北京茶月饼	Beijing tea mooncake
1028433974716967148	1975	金凤呈祥	Jinfengchengxiang
1028433974716967148	1975	金凤呈祥200	Jinfengchengxiang 200
1028433974716967148	1976	美优食品	Meiyou food
1028433974716967148	1977	XQB38-83皮带	XQB38-83 belt
1028433974716967148	1978	味多美卡	Meiduomei gift card
1028433974716967148	1978	Laver丰胸精油	Laver breast augmentation oil
1028433974716967148	1978	AOC 拉莫圣日尔曼干红葡萄酒 750ml	AOC Saint Germain Rameau claret 750ml
1028433974716967148	1978	AOC银奖圣玛杰庄园干红葡萄酒 750ml	AOC silver award Domaine Saint Majan claret 750ml
1028433974716967148	1978	圣玛杰庄园干红葡萄酒 750ml	Domaine Saint Majan claret 750ml
1028433974716967148	1978	红绳	Red rope
1028433974716967148	1978	红绳批发	Red rope wholesale
1028433974716967148	1978	项链挂绳编织	Necklace rope woven

Average number of queries per session and per task.

Session type	Number of queries per session	Number of queries per task
One task	1.45	1.45
Two tasks	2.93	1.47
Three or more tasks	6.14	1.43

Average duration of tasks.

Session type	Average duration of tasks
One task	37 minutes 10 seconds
Two tasks	27 minutes 26 seconds
Three or more tasks	19 minutes 44 seconds

Task identification results.

Method	Dictionary	Threshold	R	P	F
Rule based	Default	0.3	0.8995	0.8542	0.8763
Rule based	Default	0.35	0.8670	0.8946	0.8806
Rule based	Default	0.4	0.8414	0.9232	0.8804
Rule based	Default + Pro-Catalog	0.3	0.8837	0.8552	0.8692
Rule based	Default + Pro-Catalog	0.35	0.8453	0.8926	0.8683
Rule based	Default + Pro-Catalog	0.4	0.8266	0.9054	0.8642
Clustering-avg	Default	0.3	0.8394	0.9192	0.8775
Clustering-avg	Default	0.35	0.8079	0.9379	0.8681
Clustering-avg	Default	0.4	0.7724	0.9527	0.8531
Clustering-avg	Default + Pro-Catalog	0.3	0.8246	0.9232	0.8711
Clustering-avg	Default + Pro-Catalog	0.35	0.7892	0.9379	0.8571
Clustering-avg	Default + Pro-Catalog	0.4	0.7557	0.9537	0.8432
Clustering-max	Default	0.3	0.9123	0.8384	0.8738
Clustering-max	Default	0.35	0.8867	0.8778	0.8822 This approach yields the highest F-score and is used to perform task identification for the rest of the dataset.
Clustering-max	Default	0.4	0.8512	0.9123	0.8807
Clustering-max	Default + Pro-Catalog	0.3	0.9015	0.8433	0.8714
Clustering-max	Default + Pro-Catalog	0.35	0.8650	0.8788	0.8719
Clustering-max	Default + Pro-Catalog	0.4	0.8138	0.9143	0.8611

Three unrelated search tasks.

SID	Time	Query terms	Query terms (translation)
13527	2013-05-29 18:27:16	海狮油	Sea lions oil
13527	2013-05-29 18:35:24	上海移动100元快充	Shanghai Mobile 100 yuan recharge
13527	2013-05-29 18:35:48	上海移动10元	Shanghai Mobile 10 yuan
13527	2013-05-29 18:35:58	上海移动100元	Shanghai Mobile 100 yuan
13527	2013-05-29 19:05:44	澄糖滋润护唇膏玫瑰粉红	Sugar moist lip balm rose pink

Session with unrelated search tasks.

SID	Time	Query terms	Query terms (translation)
67804	2013-05-17 10:51:33	内存卡16g正品包邮	Memory card 16g free delivery
67804	2013-05-17 10:53:35	vip16g正品包邮	Vip memory card 16g free delivery
67804	2013-05-17 11:12:08	水杯	Water cup

Session duration.

Item	Session duration
Average session duration	49 minutes 3 seconds
Average task duration	27 minutes 36 seconds
Longest session	14 hours 56 minutes 22 seconds

eISSN:: 2543-683X
Language:: English

Publication timeframe:: 4 times per year
Journal Subjects:: Computer Sciences, Information Technology, Project Management, Databases and Data Mining

Journal RSS Feed

Identification and Analysis of Multi-tasking Product Information Search Sessions with Query Logs

Article Category: Research Paper

Published Online: Sep 01, 2017

Page range: 79 - 94

Received: Mar 16, 2016

Accepted: Jun 06, 2016

DOI: https://doi.org/10.20309/jdis.201621

Keywords
Product search, Shopping task identification, Shopping task analysis, Multitasking session

© 2016 Xiang Zhou, Pengyi Zhang, Jun Wang

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Figure 1

Figure 2

Figure 3

Session with related search tasks.

Distribution of the sessions according to the number of task per session.

Three related search tasks.

Average session duration.

Average query length.

Sample of human task identifications.

Basic characteristics of sessions and tasks.

Sample query records.

Average number of queries per session and per task.

Average duration of tasks.

Task identification results.

Three unrelated search tasks.

Session with unrelated search tasks.

Session duration.

Identification and Analysis of Multi-tasking Product Information Search Sessions with Query Logs

Article Category: Research Paper

Published Online: Sep 01, 2017

Page range: 79 - 94

Received: Mar 16, 2016

Accepted: Jun 06, 2016

DOI: https://doi.org/10.20309/jdis.201621

KeywordsProduct search, Shopping task identification, Shopping task analysis, Multitasking session

© 2016 Xiang Zhou, Pengyi Zhang, Jun Wang

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Figure 1

Figure 2

Figure 3

Session with related search tasks.

Distribution of the sessions according to the number of task per session.

Three related search tasks.

Average session duration.

Average query length.

Sample of human task identifications.

Basic characteristics of sessions and tasks.

Sample query records.

Average number of queries per session and per task.

Average duration of tasks.

Task identification results.

Three unrelated search tasks.

Session with unrelated search tasks.

Session duration.

Keywords
Product search, Shopping task identification, Shopping task analysis, Multitasking session