We measure their performance on the cold-start users and all results are shown as Fig. 1. Obviously, the User-based CF fails to recommend with 15.73% coverage. The reason is that the cold-start users rate few services, so it is difficult to compute the similarities. After introducing the trust relationships, the Coverage of TidalTrust and MoleTrust is bigger than User-based CF. At the same time, the RMSE decrease. But the precision does not improve obviously. The trust-related algorithms based on random walk model gain the significant improve on the coverage and precision according to the users’ selection strategy. And the RMSE of TSWSWalker is the smallest because the selected users of one walk are the most trustful and similar with the source user. After analyzing the data, there are about 26% users from the selected users in our algorithm are not considered in the RelevantTrustWalker algorithm. Finally, the F-measure of the TSWSWalker is better than the other algorithms.
A Novel Service Recommendation Approach 435
Fig. 1. Comparison results for cold start users
From the Fig. 2, we can see that all the trust-related algorithms outperform the User- based CF for all considered metrics. For the algorithms based on the random walk, it is better to design the user-selection strategy in the walk process, i.e., the RMSE of TSWSWalker and RelevantTrustWalker are less than TrustWalker’s. Even though the
Fig. 2. Comparison results for all users.
436 G. Li et al.
coverage of TSWSWalker and RelevantTrustWalker are very close, TSWSWalker’s RMSE is smaller than RelevantTrustWalker. The reason is that there are about 89% users who have common trusted users with others after analyzing the dataset. In the same way, the recommendation performance of the TSWSWalker outperforms the other algorithms for all users.
5 Conclusion
In this paper, an improved random walk algorithm is proposed to recommend the serv‐
ices. The non-negative matrix factorization technique is hired to compute the similarities of users and services. In each walk, we not only consider the rating similarity between users but their trust value. At last, the extensive experiments are conducted and the results validate that our proposed approach outperforms the other recommendation algorithms.
In the used dataset, there exist distrust relationships. These distrust users’ influence on the recommendation accuracy will be deeply studied as a future work. And we will compare more algorithms with our algorithm.
Acknowledgments. This work was funded by the Natural Science Foundation of Shandong Province (NSFS Grant No. ZR2014FL013) and the Independent Innovation and Achievements Transformation Special Project of Shandong Province (No. 2014ZZCX02702). The authors acknowledge the support of the Opening Fund of Shandong Provincial Key Laboratory for Network Based Intelligent Computing.
References
1. Al-Masri, E., Mahmoud, Q.H.: QoS-based discovery and ranking of web services. In: The 16th International Conference on Computer Communications and Networks, pp. 529–534.
IEEE Press, Honolulu, Hawaii (2007)
2. Zheng, Z., Ma, H., Lyu, M.R., King, I.: WSRec: a collaborative filtering based web service recommender system. In: The 16th International Conference on Web Services, pp. 437–444.
IEEE Computer Society, Los Angeles (2009)
3. Massa, P., Avesani, P.: Trust-aware collaborative filtering for recommender systems. In:
Meersman, R., Tari, Z. (eds.) OTM 2004. LNCS, vol. 3290, pp. 492–508. Springer, Heidelberg (2004). doi:10.1007/978-3-540-30468-5_31
4. Wang, S., Hsu, C.-H., Liang, Z., Sun, Q.: Multi-user web service selection based on multi- QoS prediction. Inf. Syst. Front. 16(1), 143–152 (2014)
5. Chen, X., Zheng, Z., Yu, Q., Lyu, M.R.: Web service recommendation via exploiting location and QoS information. IEEE Trans. Parallel Distrib. Syst. 25(7), 1913–1924 (2014)
6. He, P., Zhu, J., Zheng, Z., Xu, J., Lyu, M.R.: Location-based hierarchical matrix factorization for web service recommendation. In: The 21st International Conference on Web Services, pp.
297–304. IEEE Computer Society, Alaska (2014)
7. Zheng, Z., Ma, H., Lyu, M.R., King, I.: Collaborative web service QoS prediction via neighborhood integrated matrix factorization. IEEE Trans. Serv. Comput. 6(3), 289–299 (2013)
A Novel Service Recommendation Approach 437
8. Yu, Q., Zheng, Z., Wang, H.: Trace norm regularized matrix factorization for service recommendation. In: 20th IEEE International Conference on Web Services, pp. 34–41. IEEE Computer Society, Santa Clara (2013)
9. Li, Z., Cao, J., Gu, Q.: Temporal-aware QoS-based service recommendation using tensor decomposition. J. Web Serv. Res. 12(1), 62–74 (2015)
10. Zhang, R., Li, C., Sun, H., Wang, Y., Huai, J.: Quality of web service prediction by collective matrix factorization. In: 11th International Conference on Service Computing, pp. 432–439.
IEEE Xplore Press, Bangalore (2014)
11. Abdullah, A.: An integrated-model QoS-based graph for web service recommendation. In:
22nd International Conference on Web Services, pp. 416–423. IEEE Computer Society, New York (2015)
12. Golbeck, J.A.: Computing and applying trust in web-based socail networks, University of Maryland (2005)
13. Dongyan, J., Fuzhi, Z.: A collaborative filtering recommendation algorithm based on double neighbor choosing strategy. J. Comput. Res. Dev. 50(5), 1076–1084 (2013)
14. He, J., Chu, W.W.: Social networ-based recommender system (SNRS). In: Memon, N., Xu, J.J., Hicks, D.L., Chen, H. (eds.) Data Mining for Social Network Data. Annals of Information Systems, vol. 12, pp. 47–74. Springer, Heidelberg (2010)
15. Ray, S., Mahanti, A.: Improving prediction accuracy in trust-aware recommender systems.
In: 43rd International Conference on System Sciences, pp. 1–9. IEEE Computer Society, New York (2010)
16. Tang, M., Xu, Y., Liu, J., Zheng, Z., Liu, X.F.: Trust-aware service recommendation via exploiting social networks. In: 10th IEEE International Conference on Services Computing, pp. 376–383. IEEE Computer Society, Santa Clara (2013)
17. Jamali, M., Ester, M.: TrustWalker: a random walk model for combining trust-based and item- based recommendation. In: 15th International Conference on Knowledge Discovery and Data Mining, pp. 397–406. ACM, Paris, France (2009)
18. Deng, S., Huang, L., Xu, G.: Social network-based service recommendation with trust enhancement. Expert Syst. Appl. 41(18), 8075–8084 (2014)
19. Tang, M., Dai, X., Cao, B., Liu, J.: WSWalker: a random walk method for Qos-aware web service recommendation. In: 22th International Conference on Web Services, pp. 591–598.
IEEE Computer Society, New York (2015)
20. Hoyer, P.O.: Non-negative matrix factorization with sparseness constraints. J. Mach. Learn.
Res. 5, 1457–1469 (2004)
21. Victor, P., Cornelis, C., De Cock, M., Teredesai, A.: Trust-and distrust-based recommendations for controversial reviews. IEEE Intell. Syst. 26(1), 48–55 (2011) 22. Massa, P., Avesani, P.: Trust-aware recommender systems. In: 1st Conference on
Recommender Systems, pp. 17–24. ACM (2007) 438 G. Li et al.
3-D Design Review System in Collaborative Design of Process Plant
Jian Zhou1(✉), Linfeng Liu1, Yunyun Wang1, Fu Xiao1, and Weiqing Tang2
1 College of Computer, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
zhoujian@njupt.edu.cn
2 Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, China
Abstract. Design review is important in collaborative design of process plants.
To satisfy the actual work demands of design review, a 3-d design review system is developed and the key technologies such as information organization model and multi-resolution rendering approach are proposed. The information organi‐
zation model combining scene tree and attribute tree can organize the information from different CAD systems with a unified structure, and optimize the information query speed. The multi-resolution rendering approach based on programmable graphics pipeline can improve rendering efficiency within less preprocessing time, without using extra hard-disk space. Examples show that the 3-d design review system can work on a general PC to review a large quantity of design information from different subjects, and ensure real-time interaction at the same time.
Keywords: Design review ã Information organization model ã Multi-resolution rendering ã Collaborative design
1 Introduction
Process plants, such as refineries and petrochemical plants, are complex facilities mainly consisting of pipelines and equipment [1]. As shown in Fig. 1, process plants are used in industries such as petrochemical, power, metallurgical industries. With increasing product complexity and intensive global competition in the process plant industry, companies are increasingly relying on collaborative design techniques to shorten the design cycle and to sustain the optimum productivity [2].
In collaborative design of process plants, there are constraints among stages or subjects. The relevant design must meet the constraints, otherwise there will be conflic‐
tion. So design review is important in collaborative design of process plants. A reviewer has to check the results of different stages or subjects to find the design errors and conflicts, and then inform relevant designers the review results. When the review effi‐
ciency is improved, the rework in construction and the corresponding cost waste can be reduced, which helps to avoid the extension of period. So far, many review systems have
© ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2017 S. Wang and A. Zhou (Eds.): CollaborateCom 2016, LNICST 201, pp. 439–450, 2017.
DOI: 10.1007/978-3-319-59288-6_40
been developed by major CAD companies for their own CAD products, e.g. Navis‐
Works, SmartPlant Review and PDMS Review. But none of them works well in large CAD datasets especially on current desktop PCs.
A
Urumqi Petrochemical Company Refinery 110 million tons / year delayed coking unit expansion project
B
Wastewater treatment project in Shanghai Baoshan Iron and Steel Dedigned by South Metallurgical Engineering design Company
Fig. 1. Instances of process plants
First, the design institutes adopt different CAD systems, leading to the heterogeneity of design information. Second, limited by economic conditions, the design institutes usually work with general PC. Finally, as the collaborative design technology is getting more widely used in design of large-scale process plants, the quantity and complexity of information in design review have been raised rapidly [1]. To solve the problems mentioned above, a new 3-d design review system needs to be developed, which can process large quantity of design information from different CAD systems while working on a general PC.
The remainder of this paper is structured as follows. We introduce the problems and some related works in Sect. 2. The architecture of our review system is described in Sect. 3. In Sect. 4, the key technologies such as information organization model and multi-resolution rendering approach are proposed. Section 5 presents and discusses the function and the performance of our review system. Finally, conclusions are drawn in Sect. 6.
2 Problems and Related Work
During design review, the reviewer could find the design errors and conflicts among subjects or stages by real-time 3-d navigation, either by referring to the attributes and design conditions, or through automatic collision check and design condition check by the computer. To achieve the above functions, some technologies must be improved and adopted in review system, e.g., fast rendering, human-computer interaction, information organization and collision detection. In this paper, we focus on two problems, how to organize the design information from different CAD systems, and how to fast render a large-scale process plant model.
440 J. Zhou et al.