Volume 23, Issue 1

Due to the currently insufficient consideration of task fitness and task coordination for task allocation in collaborative customized product development, this research was conducted based on the analysis of collaborative customized product development process and task allocation strategy. The definitions and calculation formulas of task fitness and task coordination efficiency were derived, and a multiobjective optimization model of product customization task allocation was constructed. A solution based on adaptive genetic algorithm was proposed, and the feasibility and effectiveness of the task allocation algorithm were tested and verified using a 5-MW wind turbine product development project as example.

The financial market is a highly complex and dynamic system that has great commercial value; thus, many financial elite are drawn to research on the subject. Recent studies show that machine learning methods perform better than traditional statistical ones. In our study, based on the characteristics of financial sequence data, we propose a wrapped ensemble approach using a supervised learning algorithm to predict stock price volatility of China’s stock markets. To check our new approach, we developed an intelligent financial forecast system and used the Hushen 300 index data to test our model; it proves that our model performs better than a single algorithm. We also compared our model with the famous ensemble approach of bagging, and the result shows that our model is better.

We propose in this article an indexing and retrieval approach applied on outline shapes. Models of objects are stored in a database using the textual descriptors of their silhouettes. We extract from the textual description a set of efficient similarity measures to index the silhouettes. The extracted features are the geometric quasi-invariants that vary slightly with the small change in the viewpoint. We use a textual description and quasi-invariant features to minimize the storage space and to achieve an efficient indexing process. We also use the quad-tree structure to improve processing time during indexing. Using both geometric features and quad-tree decomposition facilitates recognition and retrieval processes. Our approach is applied on the outline shapes of three-dimensional objects. Experiments conducted on two well-known databases show the efficiency of our method in real-world applications, especially for image indexing and retrieval.

With the increasing economic development and gradual opening of China’s aviation market, the Chinese aviation industry is experiencing rapid growth and is facing some problems at the same time. Many scholars have assessed the profitability of the Chinese aviation industry using a single assessment approach. Both theory and methods in assessing the profitability of the airlines need to be developed and improved. This article presents a new integrated approach – a combination of factor analysis, principal component analysis, and gray relational analysis. Based on the theory of decision making and data mining, this more reasonable and effective model overcomes the one-sidedness and limitations of the individual methods and has significant impact in theory and practice.

Traditional document clustering algorithms consider text-based features such as unique word count, concept count, etc. to cluster documents. Meanwhile, event mining is the extraction of specific events, their related sub-events, and the associated semantic relations from documents. This work discusses an approach to event mining through clustering. The Universal Networking Language (UNL)-based subgraph, a semantic representation of the document, is used as the input for clustering. Our research focuses on exploring the use of three different feature sets for event clustering and comparing the approaches used for specific event mining. In our previous work, the clustering algorithm used UNL-based event semantics to represent event context for clustering. However, this approach resulted in different events with similar semantics being clustered together. Hence, instead of considering only UNL event semantics, we considered assigning additional weights to similarity between event contexts with event-related attributes such as time, place, and persons. Although we get specific events in a single cluster, sub-events related to the specific events are not necessarily in a single cluster. Therefore, to improve our cluster efficiency, connective terms between two sentences and their representation as UNL subgraphs were also considered for similarity determination. By combining UNL semantics, event-specific arguments similarity, and connective term concepts between sentences, we were able to obtain clusters for specific events and their sub-events. We have used 112 000 Tamil documents from the Forum for Information Retrieval Evaluation data corpus and achieved good results. We have also compared our approach with the previous state-of-the-art approach for Router-RCV1 corpus and achieved 30% improvements in precision.

Predicting the defect-prone modules when the previous defect labels of modules are limited is a challenging problem encountered in the software industry. Supervised classification approaches cannot build high-performance prediction models with few defect data, leading to the need for new methods, techniques, and tools. One solution is to combine labeled data points with unlabeled data points during learning phase. Semi-supervised classification methods use not only labeled data points but also unlabeled ones to improve the generalization capability. In this study, we evaluated four semi-supervised classification methods for semi-supervised defect prediction. Low-density separation (LDS), support vector machine (SVM), expectation-maximization (EM-SEMI), and class mass normalization (CMN) methods have been investigated on NASA data sets, which are CM1, KC1, KC2, and PC1. Experimental results showed that SVM and LDS algorithms outperform CMN and EM-SEMI algorithms. In addition, LDS algorithm performs much better than SVM when the data set is large. In this study, the LDS-based prediction approach is suggested for software defect prediction when there are limited fault data.

A new medical image enhancement algorithm based on spatial frequency domain is presented in this article. The medical image is first divided into several sub-images based on dyadic wavelet scale analysis. At each level, different directional sub-band images can reflect the different characteristics of the image. A low-frequency sub-band image maintains the original image content information, and high-frequency sub-band images represent image details such as edges and regional boundaries. The corresponding sub-band images are then enhanced by different Butterworth homomorphic filtering functions, which can attenuate the low frequencies and amplify the high frequencies. A linear adjustment is carried out on the low frequency of the highest level. Then, the wavelet reconstruction course is used to obtain the final enhanced image. Experiments on magnetic resonance images of temporomandibular joint soft tissues have shown that the proposed method can effectively eliminate the non-uniform luminance distribution of medical images. Its performance is much better than traditional Butterworth homomorphic filtering algorithm whether in subjective vision quality or objective evaluations such as detailed information entropy and average gradient.

In a distributed multi-agent system, an idle agent may be available to assist other agents in the system. An agent architecture called intent recognition is proposed in this article to accomplish this with minimal communication. To assist other agents in the system, an agent performing recognition observes the tasks other agents are performing. Unlike the much-studied field of plan recognition, the overall intent of an agent is recognized instead of a specific plan. The observing agent may use capabilities that it has not observed. In this study, the key research question is: What are intent-recognition systems and how can these be used to have agents autonomously assist each other effectively and efficiently? A conceptual framework is proposed to address this question. An implementation of the conceptual framework is tested and evaluated. A set of metrics, including task time and number of communications, is used to compare the performance of plan recognition and intent recognition. This research shows that under certain conditions, an intent-recognition system is more efficient than a plan recognition system.