Catálogo de publicaciones - libros

This study presents a novel model that integrates agent negotiation into adaptive learning for enhancing interaction efficiency between learner and instructor and promoting learning effectiveness. A constraint-based agent negotiation mechanism is employed to support a one-to-one interaction. Through the negotiation process, the instructor also can gradually perceive the learners’ feedback and then reflect on the appropriateness of the learning sequence, adjust instructional goal, approach and scheme. The instructor can thus provide more adaptive teaching based on learners’ specific needs to enhance learning effectiveness. Experimental results suggested that the proposed methodology was able to improve learning performance and learners also believed that the system enhanced their learning motivation and increased the flexibility of course learning.

The recent rush towards electronic commerce over the Internet raises many challenges, this paper present a simple model of supply chains. First a model of supply chains is discussed, highlighting two characteristic features: hierarchical subtask decomposition, and resource contention. Then a market protocol based on distributed but concurrent auctions is proposed. The protocol allow each of these markets to function separately, while the information exchanged between a sequence of markets along a single supply chain to ensure efficient global behavior across the supply chain. Each market that forms a link in the supply chain operates as a double auction, where the bids on one side of the double auction come from bidders in the corresponding segment of the industry, and the bids on the other side are synthetically generated by the protocol to express the combined information from all other links in the chain.

This paper proposes an enhanced method of Web information extraction by exploiting general phenomena that Web pages in a site tend to have common structures and dynamic Web pages contain multiple data blocks with repeating structural patterns. By considering this kind of regularity in dynamic Web pages, we develop a data block extraction system which basically adopts a supervised learning mechanism with training and extraction phases. In the training phase, the user selects and specifies a data block and the extraction rules for the block are generated. During this phase, the block is defined with the HTML DOM-tree path to the block and the tag sequence of the block. In the extraction phase, the rules are applied to the target pages to extract those blocks that have similar structure as the user-defined block. A series of experiments are performed to evaluate the user-defined data block extraction method for a number of well-known Web sites with dynamic Web pages, and the result of evaluation is satisfactory with high precision and recall measures.

This paper explores the attributes that drive company wealth creation in the Miscellaneous Industrials sector of the Australian Stock Market. It looks at how the company’s wealth creation changes in comparison to the changes in the Miscellaneous Industrial Index. We examine traditional and artificial intelligent (AI) feature selection techniques, to select attributes that drive company wealth and observe if a multiple domain model outperforms a single domain model with regards to predicting company wealth. Using a large number of calculated attributes, our empirical findings suggest that a multiple domain model was most effective. We found that WACC, Funds from Operation / EBITDA and EPS assist in guiding the direction of change in shareholder wealth. Whereas ROA, Capital Turnover and Gross Debt / Cashflow are key attributes in understanding the behaviour of the relative shareholder growth. We observed that ROIC, Ordinary Share Price, EVA, EPS and Trading Revenue / Total Assets are the important attributes that drive relative shareholder wealth in this industry.

The real-time information access of goods plays an increasingly important role for decision-making in operations management of supply chain. Based on mobile agent and radio-frequency identification (RFID) technology, this paper proposes a novel solution to track in transit goods. Firstly, mobile RFID reader integrated in a smart phone retrieves information automatically from RFID tags bound to goods. Then, the front-end system that works on the smart phone send information back to the back-end system via short message system of mobile telecommunication platform. Because the solution is built on mobile agent platform, it enables the back-end system to dispatch mobile agents to the front-end system to process raw RFID data locally, and only the filtered results are fed back. Therefore, the solution promises to reduce time and cost of data transfer between the front-end system and the back-end system. Finally, a prototype system, GTSys named, is implemented and verified on simulation platform.

Model-based diagnosis is one of the active branches of Artificial Intellgence. Conflict recognition, aiming at generating all minimal conflict sets (MCSs), and candidate generation, aiming at generating all minimal hitting sets (MHSs), are of the two important steps towards to the final diagnosis results. Firstly an SE-tree based algorithm (CSSE-tree) for deriving all MCSs is given. Then a concept of inverse SE-tree (ISE-tree) is put forward, and an ISE-tree based algorithm (CSISE-tree) for deriving all MCSs is presented as well. Considering the similarity of generation of all MCSs and all MHSs for the collection of all MCSs, a uniform framework for deriving all MCSs and MHSs is proposed, too. Experimental results show that our algorithms have better efficiency than others in some situations.

This work adopted the fuzzy constraint-directed approach to model opponent’s beliefs in agent negotiation. The fuzzy constraint-directed approach involves the fuzzy probability constraint and the fuzzy instance reasoning. The fuzzy probability constraint is used to cluster the opponent’s regularities and to eliminate the noisy hypotheses or beliefs, so as to increase the efficiency on the convergence of behavior patterns and to improve the effectiveness on beliefs learning. The fuzzy instance reasoning reuses the prior opponent knowledge to speed up problem-solving, and reason the proximate regularities to acquire desirable results on predicting opponent behavior. Besides, the proposed interaction method allows the agent to make a concession dynamically based on desirable objectives. Moreover, experimental results suggest that the proposed framework enabled an agent to achieve a higher reward, a fairer deal, or a less cost of negotiation.

Software agents representing supply chain partners make it possible to automate supply chain management and particularly can address the challenging problem of automating the process of dynamic supply chain formation. This paper puts forward a multi-agent negotiation mechanism for dynamic supply chain formation in semi-monopolized circumstance, i.e., China petroleum supply chain, where the conventional negotiation mechanisms are limited because they are based on the assumption of a pure market. The proposed multi-agent negotiation mechanism is algorithmized and validated, respectively.

Web-learning based on multi-agents is a hotspot in the field of computer software research; this paper aims to present an interactive and collaborative learning environment, a new type of web-learning system, designed for educational purpose and gives a brief summary of agent and creates the structure model of the intelligent web-learning based on multi-agents according to the theory of Constructivism, and it also discusses the function and the structure of every agent. It also discusses that how the functions are carried out.

The issue of coalition formation problem has been investigated from many aspects. However, all of the previous work just take the capability of agent into account, but not consider those factors, such as the time that agent takes to achieve a task, the cost of employing agent, the credit standing of agent, the risk that the task sponsor bears, and the bias of task sponsor and so on. So we originally take these factors into account. The coalition problem in this paper is a constrained problem including a great deal of equality constraints and inequality constraints. So we adopt the to transform it to an unconstrained one. That is to say, it becomes a single objective function. Being an unconstrained optimization algorithm, the binary particle swarm optimization algorithm is adopted to address this problem. To improve the capability of global searching of our algorithm and convergent rate of the solutions, we divide the process of coalition formation into two stages to deal with respectively. Simulations show that our algorithm is effective and feasible.