Dynamic Pricing in Spatial Crowdsourcing：A Matching-Based Approach

Dynamic Pricing in Spatial Crowdsourcing A Matching-Based Approach Lei Chen 3, Bowen Du, Jieping Yer x Yongxin Tong, Libin Wang, Zimu zhou 1 Beihang University Z ETH Zurich 3 The Hong Kong University of Science and Technology 4 DiDi Chuxing Inc ⑦其章航家[Enh西删 DiDi More than a jounsy

Dynamic Pricing in Spatial Crowdsourcing: A Matching-Based Approach Yongxin Tong1 , Libin Wang1 , Zimu Zhou2 , Lei Chen3 , Bowen Du1 , Jieping Ye4 1 Beihang University 2 ETH Zurich 3 The Hong Kong University of Science and Technology 4 DiDi Chuxing Inc

Outline o Background and Motivation ● Problem statement ° Our so| utions ● EXperiments Conclusion

Outline ⚫ Background and Motivation ⚫ Problem Statement ⚫ Our Solutions ⚫ Experiments ⚫ Conclusion

Spatial Crowdsourcing A new way of organizing the crowd(mobile Internet workers) to do spatial tasks Representatives O DiDi uber Gigwall GRUBHUB Y Tong, L Chen, and C Shahabi Spatial Crowdsourcing: Challenges, Techniques, and Applications. (tutorial) VLDB 2017

⚫ A new way of organizing the crowd (mobile Internet workers) to do spatial tasks ⚫ Representatives Spatial Crowdsourcing Y. Tong, L. Chen, and C. Shahabi. Spatial Crowdsourcing: Challenges, Techniques, and Applications. (tutorial) VLDB 2017

Spatial crowdsourcing: Workflow Confirm Request Express Luxe Premia Requester (1) Request (3 )Decision e Didi Chux It serves as an important step Board after minls Patm"意(2) Pricing (4)Assignment Beihang U niversity Comfort Worker approx. 51. 9cNY Request a premier car

Spatial Crowdsourcing: Workflow Requester Platform Worker (1) Request (2) Pricing (3) Decision (4) Assignment It serves as an important step

Outline e background and motivation ● Problem statement ° Our so| utions ● EXperiments Conclusion

Outline ⚫ Background and Motivation ⚫ Problem Statement ⚫ Our Solutions ⚫ Experiments ⚫ Conclusion

Basic Concepts Requester 1 T3. A set of requesters/tasks R, for Trip distance dr(km)1.3.7 1 achr∈R ● origin or destination des private valuation vr($/km) accepts a unit price p if vr>p r2(26 k(15)wr(3,5) W2C5) A set of workers w for eachwew ●| ocation l 3653) adis a 01234 678

1 2 3 4 5 1 2 3 4 5 0 X Y 6 7 8 6 7 8 ⚫ A set of requesters/tasks 𝑹 , for each 𝒓 ∈ 𝑹: ⚫ origin 𝒐𝒓𝒊𝒓 ⚫ destination 𝒅𝒆𝒔𝒓 ⚫ private valuation 𝒗𝒓 （$/km） o accepts a unit price 𝒑 if 𝒗𝒓 > 𝒑 ⚫ A set of workers 𝑾, for each 𝒘 ∈ 𝑾: ⚫ location 𝒍𝒘 ⚫ radius 𝒂𝒘 𝒓𝟐(𝟐,𝟔) 𝒓𝟑(𝟓,𝟓) Requester 𝒓𝟏 𝒓𝟐 𝒓𝟑 Trip distance 𝒅𝒓 (km) 1.3 0.7 1 𝒘𝟏(𝟑,𝟓) 𝒘𝟐(𝟕,𝟓) 𝒘𝟑(𝟓,𝟑) 𝒓𝟏(𝟏,𝟓) Basic Concepts

Basic Concepts Requester 1T2T3 Goal: platforms set prices to Trip distance d,(km)1.3.7 1 maximize the total revenue! The revenue of request r is drp Platform sets the same (unit) price p for requesters in the same grid g 15 16 U BER 6 r3(55 r015)w735 w2(,5) 3) 01234 678X

1 2 3 4 5 1 2 3 4 5 0 X Y 6 7 8 6 7 8 Basic Concepts 𝒓𝟏(𝟏,𝟓) 𝒓𝟐(𝟐,𝟔) 𝒓𝟑(𝟓,𝟓) ⚫ The revenue of request 𝒓 is 𝒅𝒓𝒑 ⚫ Platform sets the same (unit) price 𝒑 for requesters in the same grid 𝒈 𝒘𝟏(𝟑,𝟓) 𝒘𝟐(𝟕,𝟓) 𝒘𝟑(𝟓,𝟑) 𝟏 𝟓 𝟗 𝟏𝟑 𝟐 𝟑 𝟒 𝟔 𝟕 𝟖 10 𝟏𝟏 𝟏𝟐 14 𝟏5 𝟏6 Goal: platforms set prices to maximize the total revenue! Requester 𝒓𝟏 𝒓𝟐 𝒓𝟑 Trip distance 𝒅𝒓 (km) 1.3 0.7 1

Basic Concepts Requester 1T2T3 Goal: platforms set prices to Trip distance d,(km)1.3.7 1 maximize the total revenue! i Platform i $3/km r(15)wr35) W2(5) 3) 012345678X

Platform 1 2 3 4 5 1 2 3 4 5 0 X Y 6 7 8 6 7 8 Basic Concepts 𝒓𝟏(𝟏,𝟓) 𝒓𝟐(𝟐,𝟔) 𝒓𝟑(𝟓,𝟓) 𝒘𝟏(𝟑,𝟓) 𝒘𝟐(𝟕,𝟓) 𝒘𝟑(𝟓,𝟑) 𝑟1 𝑟2 𝑟3 𝑤1 𝑤2 𝑤3 $3/km Goal: platforms set prices to maximize the total revenue! Requester 𝒓𝟏 𝒓𝟐 𝒓𝟑 Trip distance 𝒅𝒓 (km) 1.3 0.7 1

Basic Concepts Requester 1T2T3 Goal: platforms set prices to Trip distance d,(km)1.3.7 1 maximize the total revenue! i Platform I $3/km 2(6) (5 r4(15 3,5 $2/km 3s53) 012345678X

Platform 1 2 3 4 5 1 2 3 4 5 0 X Y 6 7 8 6 7 8 Basic Concepts 𝒓𝟏(𝟏,𝟓) 𝒓𝟐(𝟐,𝟔) 𝒓𝟑(𝟓,𝟓) 𝒘𝟏(𝟑,𝟓) 𝒘𝟐(𝟕,𝟓) 𝒘𝟑(𝟓,𝟑) 𝑟1 𝑟2 𝑟3 𝑤1 𝑤2 𝑤3 $3/km $2/km Goal: platforms set prices to maximize the total revenue! Requester 𝒓𝟏 𝒓𝟐 𝒓𝟑 Trip distance 𝒅𝒓 (km) 1.3 0.7 1

Basic Concepts Requester 1T2T3 Goal: platforms set prices to Trip distance d,(km)1.3.7 1 maximize the total revenue! Wn,>3 71 $3/km v<3 2(6) (55 r(15)w735) W2(5) $2/km 3) Decision 01234 678X

Basic Concepts 1 2 3 4 5 1 2 3 4 5 0 X Y 6 7 8 6 7 8 𝒓𝟏(𝟏,𝟓) 𝒓𝟐(𝟐,𝟔) 𝒓𝟑(𝟓,𝟓) 𝒘𝟏(𝟑,𝟓) 𝒘𝟐(𝟕,𝟓) 𝒘𝟑(𝟓,𝟑) 𝑟1 𝑟2 𝑟3 𝑤1 𝑤2 𝑤3 $3/km $2/km Decision 𝑣𝑟1 > 3 𝑣𝑟3 > 2 Goal: platforms set prices to maximize the total revenue! 𝑣𝑟2 < 3 Requester 𝒓𝟏 𝒓𝟐 𝒓𝟑 Trip distance 𝒅𝒓 (km) 1.3 0.7 1