RECOMMENDING PRODUCTION PLANS

• US 20140297357A1
• Filed: 11/21/2011
• Published: 10/02/2014
• Est. Priority Date: 11/21/2011
• Status: Active Grant

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Zeng, Jun, Nezhad, Hamid Reza Motahari
• Granted Patent: US 10,789,561 B2
• Time in Patent Office: Days
• US Class Current: 705/7.26
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/06316 Sequencing of tasks or work
• CPC Class Codes: G06Q 50/08 Construction
• CPC Class Codes: G06Q 50/16 Real estate

Abstract

In one embodiment, a method of recommending a production plan includes calculating a similarity score between an incoming order and each historical order in a historical order database, providing a list of most similar historical orders and corresponding historical production plans ranked according to highest similarity scores, receiving an election indicating a historical production plan as a selected production plan, and admitting the selected historical production plan to fulfill the incoming order.

Claims

1. A method of recommending a production plan comprising:

calculating a similarity score between an incoming order and each historical order in a historical order database;

providing a list of most similar historical orders and corresponding historical production plans ranked according to highest similarity scores;

receiving an election indicating a historical production plan as a selected production plan; and

admitting the selected historical production plan to fulfill the incoming order.

2. A method as in claim 1, wherein calculating a similarity score between an incoming order and each historical order comprises calculating a normalized, aggregated, weighted similarity score of the incoming order against each historical order.

3. A method as in claim 1, further comprising saving the selected production plan and corresponding historical order in the database upon receiving an indication that the incoming order has been successfully filled.

4. A method as in claim 1, further comprising customizing the selected production plan before admitting the selected production plan.

5. A method as in claim 4, wherein customizing the selected production plan comprises:

receiving a request to change a step in the selected production plan;

providing a list of candidate replacement steps;

receiving an election that identifies a candidate replacement step as a replacement step; and

replacing the step in the selected production plan with the replacement step.

6. A method as in claim 5, wherein customizing the recommended historical production plan further comprises:

after replacing the step, providing a list of candidate next-steps;

receiving an election that identifies a candidate next-step as a next step; and

inserting the next step into the selected production plan following the replacement step.

7. A method of recommending a production plan comprising:

calculating a similarity score between an incoming order and each historical order in a historical order database;

determining from the similarity scores, that no production plan associated with any historical order exceeds a similarity threshold;

requesting a first step for a new production plan;

upon receiving a first step, generating a list of candidate next-steps;

upon receiving an election for a next step, inserting the next step in the new production plan and generating another list of candidate next-steps;

upon receiving no additional elections, admitting the new production plan to fulfill the incoming order.

8. A method as in claim 7, wherein generating a list of candidate next-steps comprises:

for each step in each historical order, calculating a step similarity score based on step text similarity and step sequence similarity; and

ranking the steps based on the step similarity scores.

9. A processor-readable medium storing code representing instructions that when executed by a processor cause the processor to:

calculate a similarity score between an incoming order and each historical order in a historical order database;

provide a list of most similar historical orders and corresponding historical production plans ranked according to highest similarity scores;

receive an election indicating a historical production plan as a selected production plan; and

admit the selected historical production plan to fulfill the incoming order.

10. The processor-readable medium of claim 9, wherein the similarity score is a normalized, aggregated, weighted similarity score calculated in accordance with

11. The processor-readable medium of claim 10, wherein an attribute similarity score comprises a textual attribute similarity score calculated by:

extracting keywords from the incoming order; and

applying a cosine similarity function to quantify textual similarities between the keywords and the historical orders in the database.

12. The processor-readable medium of claim 11, wherein the cosine similarity function is applied in accordance with

Cosine Similarity=(Va·

Vb)/(∥

Va∥

*∥

Vb∥

)

wherein;

(kw1, kw2, . . . , kwn) is the union of a set of keywords extracted from two textual fields a and b;

Va and Vb are vectors representing the significance of extracted keywords in textual fields a and b;

∥

Va∥ and

∥

Vb∥

represent vector magnitudes; and

Va·

Vb represents the dot product of vectors Va and Vb.

13. The processor-readable medium of claim 10, wherein an attribute similarity score comprises a categorical attribute similarity score determined by:

assigning a numerical value to each categorical value of a categorical attribute based on a predetermined impact the categorical attribute has on production planning;

calculating the absolute difference between the numerical value of an incoming order and the numerical value of a historical order.

14. The processor-readable medium of claim 9, further storing code representing instructions that when executed by the processor cause the processor to save the selected production plan in the database upon receiving an indication that the incoming order has been successfully filled.

15. The processor-readable medium of claim 9, further storing code representing instructions that when executed by the processor cause the processor to customize the selected production plan before admitting the selected production plan for fulfillment of the incoming order.

16. The processor-readable medium of claim 15, wherein customizing the selected production plan comprises:

receiving a request to change a step in the selected production plan;

providing a list of candidate replacement steps;

receiving an election that identifies a candidate replacement step as a replacement step; and

replacing the step in the selected production plan with the replacement step.

17. The processor-readable medium of claim 16, wherein customizing the recommended historical production plan further comprises:

after replacing the step, providing a list of candidate next-steps;

receiving an election identifying a candidate next-step as a next step; and

inserting the next step into the selected production plan following the replacement step.

18. The processor-readable medium of claim 16, wherein providing a list of candidate replacement steps comprises:

for each step in each historical order, calculating a step similarity score based on step text similarity and step sequence similarity; and

ranking the steps based on the step similarity scores.

19. The processor-readable medium of claim 18, wherein calculating a step similarity score comprises quantifying the similarity between two steps and assigning a most similar step as a corresponding step in accordance with

Sim(s_n,new,s_i,old)=textSim(s_n,new,s_i,old)+pathSim(s_n,new,s_i,old)

where pathSim(s_n,new,s_i,od)=Σ

₁ⁱtextSim(s_j,new,s_j,old)

wherein;

S_n,new is the last step committed in the selected production plan;

S_i,old is a step in a historical production plan currently under similarity evaluation;

textSim( ) function is a measure of the similarity of keyword text on labels of a step; and

pathSim( ) function is a measure of the similarity of the step sequence.

20. The processor-readable medium of claim 18, wherein ranking the corresponding steps comprises ranking the corresponding steps based on a weighted sum of the similarity scores between the incoming order and each historical order and similarities between the corresponding steps in accordance with

S_{j,Recommended}=Sim(O,S_j)*w₁+Sim(S_n,new,S_i,old)*w₂

wherein:

O is a current incoming order;

sn,new is the last step committed in the selected production plan;

Sj is a historical order;

si,old is a step in the historical production plan currently under similarity evaluation; and

w₁ and w₂ are weight parameters.

21. A processor-readable medium storing code representing instructions that when executed by a processor cause the processor to:

calculate a similarity score between an incoming order and each historical order in a historical order database;

determine from the similarity scores, that no production plan associated with any historical order exceeds a similarity threshold;

request a first step for a new production plan;

upon receiving a first step, generate a list of candidate next-steps;

upon receiving an election for a next step, insert the next step in the new production plan and generate another list of candidate next-steps;

upon receiving no additional elections, admit the new production plan to fulfill the incoming order.

22. The processor-readable medium of claim 21, wherein generating a list of candidate next-steps comprises:

for each step in each historical order, calculating a step similarity score based on step text similarity and step sequence similarity; and

ranking the steps based on the step similarity scores.