INFERING TRAVEL PATH IN PUBLIC TRANSPORTATION SYSTEM

• US 20140095423A1
• Filed: 10/25/2013
• Published: 04/03/2014
• Est. Priority Date: 09/29/2012
• Status: Abandoned Application

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Duan, Ning, Fei, Xiang, Gao, Peng, Hu, Ke, Wang, Zhi Hu, Zhang, Xin, Gong, Min, Qiu, Yun Jie
• Application Number:
  US14/063,731
• Publication Number:
  US 20140095423A1
• Time in Patent Office: Days
• US Class Current: 706/46
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06Q 10/025 Coordination of plural reservations, e.g. plural trip segments, transportation combined with accommodation

Abstract

A method and apparatus are disclosed for inferring a travel path in a public transportation system. The apparatus comprises: a single-line origin-destination (OD) inferring device configured to infer from boarding data of a bus line a passenger alighting stop on the line as an inferred passenger alighting stop according to historical data, wherein the boarding data comprises a passenger boarding stop during a predetermined time period and a number of boarding passengers at the stop. A transfer line inferring device is configured to infer a passenger transfer line as an inferred passenger transfer line according to the boarding data and the inferred passenger alighting stop obtained by the single-line OD inferring means.

Claims

1. An apparatus for inferring a travel path in a public transportation system, comprising:

single-line origin-destination (OD) inferring means configured to infer, from boarding data of a bus line, a passenger alighting stop on the line as an inferred passenger alighting stop according to historical data, wherein the boarding data comprises a passenger boarding stop during a predetermined time period and a number of boarding passengers at the stop; and

transfer line inferring means configured to infer a passenger transfer line as an inferred passenger transfer line according to the boarding data and the inferred passenger alighting stop obtained by the single-line OD inferring means.

2. The apparatus according to claim 1, wherein the single-line OD inferring means comprises:

an alighting probability calculator configured to calculate probabilities that the passenger alights at various stops according to passenger behavior analysis data; and

an alighting stop assigner configured to assign an alighting stop to the passenger as an inferred passenger alighting stop according to the probabilities calculated by the alighting probability calculator that the passenger alights at various stops.

3. The apparatus according to claim 2, wherein the passenger behavior analysis data comprises one or more of:

tidal passenger flow data; and

taken stop number probability distribution.

4. The apparatus according to claim 3, further comprising a weight setter, wherein the weight setter is configured to set for the alighting probability calculator a weight of the tidal passenger flow data and a weight of the taken stop number probability distribution, respectively.

5. The apparatus according to claim 1, wherein the transfer line inferring means comprises:

a transfer line probability calculator configured to calculate probabilities that the passenger transfers to various lines according to passenger behavior analysis data; and

a transfer line assigner configured to assign a transfer line to the passenger as an inferred passenger transfer line according to the probabilities calculated by the transfer line probability calculator that the passenger transfers to various lines.

6. The apparatus according to claim 5, wherein the passenger behavior analysis data comprises one or more of:

transfer angle constraint data; and

similarity constraint data.

7. The apparatus according to claim 6, further comprising a weight setter, wherein the weight setter is used for setting for the transfer line probability calculator a weight of the transfer angle constraint data and a weight of the similarity constraint data, respectively.