PAKDD'15 Data Mining Competition

9 years, 7 months ago

PAKDD'15 Data Mining Competition: Gender Prediction Based on E-commerce Data

PAKDD'15 Data Mining Competition: Gender Prediction Based on E-commerce Data is our first competition organized within the frame of The Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD). We would like to challenge participants with a task of devising effective algorithms for recognizing a gender of e-store clients. A data set for the competition was provided by FPT Group which is also the main sponsor of the awards.

Overview

FPT has always been a leading information and communication technology enterprise in Vietnam. By 2014, its revenue was about 1.65 billion US dollars, creating more than 22 thousand full-time jobs for the society. FPT has operations in 17 countries including Vietnam, Laos, Cambodia, America, Japan, Singapore, Germany, Myanmar, France, Malaysia, Australia, Thailand, United Kingdom, Philippines, Kuwait, Bangladesh and Indonesia. The main businesses are Software Development, System Integration, Information Technology Services, Distribution and Manufacturing of Information Technology products and Retails, Internet Services Providing and Data Center Services, Online News and Advertising, e-Commerce, Educational Services, Financing Services.

In e-Commerce, FPT runs several B2B2C (business-to-business-to-customer) services that provide online shopping sites and mobile applications for small and medium sellers. Transaction data, such as product browsing and purchasing activities, from buyer, and product portfolio, from seller, can be aggregated, to provide more efficient buying and selling experiences. For example, statistical machine learning techniques can be applied to predict the optimal organization and display of products that maximize the chance of bringing useful information to user, facilitate the online purchases. Perhaps, one of the vital insights, especially for fashion-related products, is the understanding of the relevancy of product to a gender of the user. In PAKDD'15 Data Mining Competition we would like to address this particular problem. More details regarding the task and a description of the competition data can be found in Task Description section.

In case of any questions please post on the forum or write us an email: son@mimuw.edu.pl

Terms & Conditions

Summary

PAKDD'15 Data Mining Competition: Gender Prediction Based on E-commerce Data has ended after over a month of intense rivalry!

The competition attracted 330 teams from which 149 participated actively by submitting at least one solution to the leaderboard. A total number of submissions was nearly 3000. From the active teams, 28 provided us a brief report describing their approach.

Awards:

The winner: Team FRDC,
Members: Ruiyu Fang (the leader), Qingliang Miao, Cuiqin Hou, Yao Meng, Lu Fang, Dajun Chen,
Affiliation: Fujitsu Research and Development Center, Beijing, China
The runner up: Team newolfy,
Members: Yingju Xia (the leader), Shuangyong Song, Qingliang Miao, Zhongquang Zheng
Affiliation: Fujitsu Research and Development Center, Beijing, China

Invitations to oral presentation at PAKDD-2015 Contest Workshop:

Team FRDC, Ruiyu Fang, Qingliang Miao, Cuiqin Hou, Yao Meng, Lu Fang, Dajun Chen (PAKDD'15 presentation)
Team newolfy: Yingju Xia, Shuangyong Song, Qingliang Miao, Zhongquang Zheng (PAKDD'15 presentation)
Team ngocan211: Pham Ngoc An, FPT University, Hanoi, Viet Nam (PAKDD'15 presentation)
Team ws: Wojciech Świeboda, University of Warsaw, Poland (PAKDD'15 presentation)
Team sohrab: Mohammad Golam Sohrab, Makoto Miwa, Yutaka Sasaki, Toyota Technological Institute, Japan (PAKDD'15 presentation)
Team kimiyoung: Zhilin Yang, Yutao Zhang, Jie Tang, Tsinghua University, Beijing, China
Team ibayer: Immanuel Bayer, University of Konstanz, Germany (PAKDD'15 presentation)
Team gambi: Maria Brbic, Dragan Gamberger, Matej Mihelcic, Matija Piskorec, Tomislav Smuc, Rudjer Boskovic Institute, Zagreb, Croatia (PAKDD'15 presentation)

I would like to thank all participants for their hard work. Congratulations on your excellent results!

In order to stimulate future research in the topic of the competition we revealed all competition data (including the labels for test cases) in the data files folder.

Terms and Conditions

The competition is open for all interested researchers, specialists and students. Only members of the Contest Organizing Committee cannot participate.
Participants may submit solutions as teams made up of one or more persons.
Each team needs to designate a leader responsible for communication with the Organizers. A single person can be a leader of only one team.
One person may be incorporated in maximally 3 teams.
Each team needs to be composed of a different set of persons.
The total number of submissions for any single team is limited to 100 solutions.
A winner of the competition is chosen on the basis of the final evaluation results. In a case of draws in the evaluation scores, time of the submission will be taken into account.
Each team is obliged to provide a short report describing their final solution. Reports must contain information such as the name of a team, names of all team members, the last preliminary evaluation score and a brief overview of the used approach. Their length should not exceed 2000 words and they should be submitted in the pdf format using our submission system by May 4, 2015. Only submissions made by teams that provided the reports will qualify for the final evaluation.
By enrolling to this competition you grant the organizers rights to process your submissions for the purpose of evaluation and post-competition research.

In case of questions related to the competition please contact us via email: webmaster@knowledgepit.fedcsis.org or through the competition forum.

Please log in to the system!

Task description

Data format: The data for participants were divided into separate training and test sets - trainingData.csv and testData.csv, respectively. Each of these files contains 15,000 records which correspond to product viewing logs. A single log is composed of four columns, separated by commas. The first one is a session ID. The second and the third column correspond to a session start time and session end time, respectively. The last column contains a list of product IDs which were viewed during the session, (the order of viewing is preserved). Consecutive product IDs are separated by semicolons. There is also available trainingLabels.csv file which contains labels identifying true gender of users whose sessions are described in the training data set.

Since a distribution of unique product IDs in the data is very sparse, the IDs contain additional information regarding product category hierarchy. Each product ID can be decomposed into four different IDs which are separated by slashes. The IDs starting with ‘A’ letter are the most general categories and those starting with ‘D’ correspond to individual products. The IDs which start with ‘B’ and ‘C’ are associated with subcategories and sub-subcategories, respectively.

Format of submissions: The participants of the competition are asked to predict the gender of users from the test data and send us their solutions using the submission system. Each solution should be sent in a single text file containing exactly 15,000 lines. The format of submitted files should follow the format of trainingLabels.csv. In the consecutive lines, this file should contain a single label which identifies the gender of a user who generated the corresponding session log in the test set.

Evaluation of results: The submitted solutions will be evaluated on-line and the preliminary results will be published on the competition leaderboard. It will correspond to approximately 20% of the test data. The final evaluation will be performed after completion of the competition using the remaining part of the test data. Those results will also be published on-line. It is important to note that only teams which submit a short report describing their approach before the end of the contest will qualify for the final evaluation. The winning teams will be officially announced during a special session at the PAKDD'15 conference (http://www.pakdd2015.jvn.edu.vn/).

Since the distribution of labels in the data is not balanced, the assessment of solutions will be done using the balanced accuracy measure which is defined as an average accuracy within the decision classes. Namely, for a vector of predictions preds and a vector of true gender labels genders we define the balance accuracy as: \[ACC_{m}(preds, genders) = \frac{|j : preds_{j} = genders_{j} = male|}{|j : genders_{j} = male|}\] \[ACC_{f}(preds, genders) = \frac{|j : preds_{j} = genders_{j} = female|}{|j : genders_{j} = female|}\] \[BAC(preds, genders) = \left(ACC_{f}(preds, genders) + ACC_{m}(preds, genders)\right)/2\]

Final results

Rank	Team Name	Is Report		Preliminary Score	Final Score	Submissions
1	ngocan211	True	True	0.8878	0.890432	2
2	frdc	True	True	0.8875	0.878928	2
3	newolfy	True	True	0.8874	0.877889	2
4	ws	True	True	0.8614	0.851119	2
5	sohrab	True	True	0.8546	0.851032	2
6	kimiyoung	True	True	0.8508	0.849046	2
7	dymitrruta	True	True	0.8474	0.847979	2
8	ibayer	True	True	0.8415	0.840673	2
9	songshuangyong	True	True	0.8534	0.830603	2
10	stderr	True	True	0.8170	0.811541	2
11	amy	True	True	0.8183	0.810673	2
12	gambi	True	True	0.8235	0.810181	2
13	siyu	True	True	0.8217	0.805283	2
14	ahwangyuwei	True	True	0.8111	0.801450	2
15	kkurach	True	True	0.8053	0.798821	2
16	duongtranduc	True	True	0.8100	0.797838	2
17	hnt	True	True	0.7949	0.797127	2
18	vinhn	True	True	0.7975	0.795656	2
19	zhu_ark	True	True	0.7948	0.791138	2
20	mrw19138	True	True	0.7947	0.790483	2
21	xiexingbo	True	True	0.7894	0.783441	2
22	pdoviet	True	True	0.7894	0.783204	2
23	jankralj	True	True	0.7849	0.780010	2
24	zhijin	True	True	0.7787	0.772296	2
25	ikttan	True	True	0.7824	0.771237	2
26	ds_tcy	True	True	0.7688	0.770413	2
27	thetuxedo	True	True	0.7764	0.766886	2
28	langochai	True	True	0.7624	0.765252	2
29	neozhangthe1	False	True	0.8530	No report file found or report rejected.	2
30	linegroup	False	True	0.8358	No report file found or report rejected.	2
31	zm	False	True	0.8216	No report file found or report rejected.	2
32	antry	False	True	0.8163	No report file found or report rejected.	2
33	mamy	False	True	0.8137	No report file found or report rejected.	2
34	tund	False	True	0.8136	No report file found or report rejected.	2
35	ggvspp	False	True	0.8135	No report file found or report rejected.	2
36	khuongnd	False	True	0.8130	No report file found or report rejected.	2
37	rjo2909	False	True	0.8115	No report file found or report rejected.	2
38	wyw	False	True	0.8114	No report file found or report rejected.	2
39	yanghaisong	False	True	0.8104	No report file found or report rejected.	2
40	thaidang	False	True	0.8091	No report file found or report rejected.	2
41	thaidt	False	True	0.8085	No report file found or report rejected.	2
42	ttd	False	True	0.8084	No report file found or report rejected.	2
43	orange	False	True	0.8046	No report file found or report rejected.	2
44	ziom	False	True	0.8040	No report file found or report rejected.	2
45	xspring	False	True	0.8066	No report file found or report rejected.	2
46	dirichlet.process	False	True	0.8020	No report file found or report rejected.	2
47	nhatuan	False	True	0.8008	No report file found or report rejected.	2
48	ramesh.krnt	False	True	0.7995	No report file found or report rejected.	2
49	ntienvu	False	True	0.8013	No report file found or report rejected.	2
50	cwhuang	False	True	0.7980	No report file found or report rejected.	2
51	fancyspeed	False	True	0.7966	No report file found or report rejected.	2
52	zhangzhongxia	False	True	0.7992	No report file found or report rejected.	2
53	derivedbydata	False	True	0.7959	No report file found or report rejected.	2
54	hnguyen	False	True	0.7957	No report file found or report rejected.	2
55	dat_phuoc	False	True	0.7952	No report file found or report rejected.	2
56	lihang00	False	True	0.7933	No report file found or report rejected.	2
57	0000	False	True	0.7930	No report file found or report rejected.	2
58	yzan	False	True	0.7920	No report file found or report rejected.	2
59	wttool	False	True	0.7919	No report file found or report rejected.	2
60	sasnzy	False	True	0.7901	No report file found or report rejected.	2
61	eric	False	True	0.7887	No report file found or report rejected.	2
62	lab213	False	True	0.7868	No report file found or report rejected.	2
63	zhili	False	True	0.7858	No report file found or report rejected.	2
64	huuphuc2609	False	True	0.7856	No report file found or report rejected.	2
65	muye5	False	True	0.7837	No report file found or report rejected.	2
66	whatif	False	True	0.7879	No report file found or report rejected.	2
67	pth1993	False	True	0.7816	No report file found or report rejected.	2
68	sslim	False	True	0.7813	No report file found or report rejected.	2
69	iran-amin	False	True	0.7801	No report file found or report rejected.	2
70	pikachust8811	False	True	0.7801	No report file found or report rejected.	2
71	binh	False	True	0.7794	No report file found or report rejected.	2
72	huong2	False	True	0.7790	No report file found or report rejected.	2
73	saj	False	True	0.7790	No report file found or report rejected.	2
74	etw	False	True	0.7788	No report file found or report rejected.	2
75	bati	False	True	0.7787	No report file found or report rejected.	2
76	ketanatnmims	False	True	0.7824	No report file found or report rejected.	2
77	p2trieu	False	True	0.7778	No report file found or report rejected.	2
78	yin520liang	False	True	0.7773	No report file found or report rejected.	2
79	tchang	False	True	0.7773	No report file found or report rejected.	2
80	baseline_solution	False	True	0.7755	No report file found or report rejected.	2
81	neurons	False	True	0.7751	No report file found or report rejected.	2
82	vutm	False	True	0.7743	No report file found or report rejected.	2
83	spiritoflinz	False	True	0.8094	No report file found or report rejected.	2
84	qjt	False	True	0.7741	No report file found or report rejected.	2
85	abhgh	False	True	0.7732	No report file found or report rejected.	2
86	wst_casd	False	True	0.7725	No report file found or report rejected.	2
87	h3p	False	True	0.7757	No report file found or report rejected.	2
88	billguess	False	True	0.7951	No report file found or report rejected.	2
89	xuxiaofeng	False	True	0.7710	No report file found or report rejected.	2
90	violet_zct	False	True	0.7765	No report file found or report rejected.	2
91	marcb	False	True	0.7693	No report file found or report rejected.	2
92	nightfury	False	True	0.7853	No report file found or report rejected.	2
93	d10207305	False	True	0.7688	No report file found or report rejected.	2
94	fajri91	False	True	0.7683	No report file found or report rejected.	2
95	mautoan11	False	True	0.7710	No report file found or report rejected.	2
96	jyclin	False	True	0.7672	No report file found or report rejected.	2
97	tidom	False	True	0.7657	No report file found or report rejected.	2
98	little_number	False	True	0.7707	No report file found or report rejected.	2
99	vpodpecan	False	True	0.7644	No report file found or report rejected.	2
100	tuandinh	False	True	0.8049	No report file found or report rejected.	2
101	gentaiscool	False	True	0.7774	No report file found or report rejected.	2
102	bruincui	False	True	0.7629	No report file found or report rejected.	2
103	starryc	False	True	0.7603	No report file found or report rejected.	2
104	jackson13	False	True	0.7602	No report file found or report rejected.	2
105	hamylinh	False	True	0.7591	No report file found or report rejected.	2
106	nampham	False	True	0.7573	No report file found or report rejected.	2
107	hoangphan	False	True	0.7653	No report file found or report rejected.	2
108	helen	False	True	0.7662	No report file found or report rejected.	2
109	kuanhoong	False	True	0.7524	No report file found or report rejected.	2
110	seeyouhere	False	True	0.7487	No report file found or report rejected.	2
111	pin	False	True	0.7455	No report file found or report rejected.	2
112	gmustafa	False	True	0.7545	No report file found or report rejected.	2
113	mayankkejriwal	False	True	0.7370	No report file found or report rejected.	2
114	mrboring	False	True	0.7365	No report file found or report rejected.	2
115	lingdian618	False	True	0.7328	No report file found or report rejected.	2
116	rembern	False	True	0.7614	No report file found or report rejected.	2
117	strnam	False	True	0.7313	No report file found or report rejected.	2
118	zagorecki	False	True	0.7844	No report file found or report rejected.	2
119	sudarsun	False	True	0.7175	No report file found or report rejected.	2
120	zgbdsg	False	True	0.7846	No report file found or report rejected.	2
121	sg.qq	False	True	0.7090	No report file found or report rejected.	2
122	sdx0112	False	True	0.7089	No report file found or report rejected.	2
123	test	False	True	0.7044	No report file found or report rejected.	2
124	blah	False	True	0.7716	No report file found or report rejected.	2
125	wilson891226	False	True	0.7897	No report file found or report rejected.	2
126	exploit	False	True	0.6976	No report file found or report rejected.	2
127	janezkranjc	False	True	0.6948	No report file found or report rejected.	2
128	chaitu516	False	True	0.6868	No report file found or report rejected.	2
129	fengqi	False	True	0.7112	No report file found or report rejected.	2
130	deagle9413	False	True	0.6796	No report file found or report rejected.	2
131	khanhlh	False	True	0.6660	No report file found or report rejected.	2
132	sathik	False	True	0.6620	No report file found or report rejected.	2
133	wolfinlove	False	True	0.6615	No report file found or report rejected.	2
134	huongtt	False	True	0.6691	No report file found or report rejected.	2
135	f-ken1010	False	True	0.6555	No report file found or report rejected.	2
136	hnt1	False	True	0.6796	No report file found or report rejected.	2
137	nghiemduc	False	True	0.7085	No report file found or report rejected.	2
138	statgeek	False	True	0.7500	No report file found or report rejected.	2
139	franksnail	False	True	0.6159	No report file found or report rejected.	2
140	meerkat	False	True	0.5641	No report file found or report rejected.	2
141	oahcil	False	True	0.5633	No report file found or report rejected.	2
142	huwenp	False	True	0.5577	No report file found or report rejected.	2
143	linjie_zhu	False	True	0.5565	No report file found or report rejected.	2
144	abhijit	False	True	0.5473	No report file found or report rejected.	2
145	ssssqd	False	True	0.5085	No report file found or report rejected.	2
146	kloud	False	True	0.5058	No report file found or report rejected.	2
147	yupbank	False	True	0.5006	No report file found or report rejected.	2
148	thnhu	False	True	0.7996	No report file found or report rejected.	2
149	customs	False	True	0.6881	No report file found or report rejected.	2
150	cllab	False	True	0.7930	No report file found or report rejected.	2
151	sink	False	True	0.5969	No report file found or report rejected.	2
152	mathimohanraj	False	True	0.0000	No report file found or report rejected.	2
153	clapika2010	False	True	0.7870	No report file found or report rejected.	2
154	pg7799	False	True	0.7766	No report file found or report rejected.	2

Schedule

March 23, 2015: start of the competition, data sets become available,
May 1, 2015: deadline for submitting the predictions,
May 3, 2015: deadline for sending the reports, end of the challenge,
May 19, 2015: beginning of the PAKDD'15 conference, official announcement of the winners.

Awards

Authors of the top ranked solutions (based on the final evaluation scores) will be awarded with prizes sponsored by FPT Group:

First Prize: Apple Mac Book Air + one free PAKDD'15 conference registration,
Second Prize: new FPT smartphone (to be defined) + one free PAKDD'15 conference registration.

The award ceremony will take place during the PAKDD'15 conference (May 19 - 22, Ho Chi Minh City, Vietnam).

Contest organizing committee

Hung Son Nguyen, University of Warsaw

Tran The Trung, FPT University

Tu Bao Ho, Japan Advanced Institute of Science and Technology

Duc Dung Nguyen, Vietnam Academy of Science and Technology

Andrzej Janusz, University of Warsaw

Forum

Discussion	Author	Replies	Last post
whether top10 reports are available sometime?	muye5	2	by Andrzej Thursday, May 21, 2015, 09:42:59
Complete test data: Gender Prediction Based on E-commerce Data	Ramesh	1	by Andrzej Thursday, May 21, 2015, 09:52:19
Postpone the deadline	Sajjad	1	by Sajjad Saturday, May 02, 2015, 00:53:18
The last week of PAKDD'15 Data Mining Competition	Andrzej	0	by Andrzej Monday, April 27, 2015, 13:06:02
Training data VS Test data	Amin	1	by Amin Sunday, April 26, 2015, 01:05:08
New competition at Knowledge Pit: IJCRS’15 Data Challenge	Andrzej	0	by Andrzej Monday, April 13, 2015, 11:22:57
A problem with KnowledgePit server	Andrzej	0	by Andrzej Sunday, April 12, 2015, 14:58:27
Final submission	Tu	2	by Andrzej Wednesday, April 01, 2015, 09:17:09
small inconsistency in test data	Vid	2	by Andrzej Monday, March 30, 2015, 10:50:10
Conference participation	Eftim	1	by Andrzej Monday, March 30, 2015, 14:03:01
AAIA'15 Data Mining Competition: Activity Recognition Based on Body Sensor Networks	Andrzej	0	by Andrzej Friday, March 27, 2015, 17:18:35
Multiple Participants in single team?	Abhay	2	by Abhay Monday, March 30, 2015, 19:36:20