Khaled El-Bahnasy

Rapid Generation Of Plant Protection Expert Systems

Annual yield losses due to pests are estimated to be in billions of dollars worldwide. Plant protection is one of the most important components of crop production in most agricultural areas of the world, and the effectiveness of crop protection depends on accurate and timely diagnosis of phytosanitary problems. A great deal of knowledge in plant protection and technologies exists in the scientific domain. The dissemination of these technologies could be enhanced by using expert systems and other artificial intelligence technologies. Expert systems are simple, yet powerful enough to provide a considerable amount of written and visual information. They are good for solving complex problems and they are modular and modifiable. They are affordable, consistent in prediction, and have the advantages of expressiveness and intuitiveness. However, a rapid generation tool is necessary to reduce the time needed to build expert systems for different crops. A domain-specific generic software for the generation of expert systems was developed, and used to build a prototype barley protection expert system based on existing knowledge. Web knowledge acquisition forms were developed for barley, chickpea, and wheat, and encoded in XML format. The knowledge bases relating to plant protection for barley, wheat, and chickpea were captured using knowledge accumulated by scientists in the co-operating centers. The knowledge base for each crop was divided into four components: variety or host plant selection, cultural practices, pest identification, and pest control.

Personalizing of Content Dissemination in Online Social Networks

Online social networks have seen a rapid growth in recent years. A key aspect of many of such networks is that they are rich in content and social interactions. Users of social networks connect with each other and forming their own communities. With the evolution of huge communities hosted by such websites, users suffer from managing information overload and it is become hard to extract useful information. Thus, users need a mechanism to filter online social streams they receive as well as enable them to interact with most similar users. In this paper, we address the problem of personalizing dissemination of relevant information in knowledge sharing social network. The proposed framework identifies the most appropriate user(s) to receive specific post by calculating similarity between target user and others. Similarity between users within OSN is calculated based on users’ social activity which is an integration of content published as well as social pattern Application of this framework to a representative subset of a large real-world social network: the user/community network of the blog service stack overflow is illustrated here. Experiments show that the proposed model outperform tradition similarity methods.

A Novel Case base Indexing Model based on Power Set Tree

CBR has been successfully applied to the areas of planning, diagnosis, law and decision making among others. It uses useful prior cases to solve the new problems. CBR must accurately retrieve similar prior cases for getting a good performance. Throughout this thesis The Novel Case Base Indexing Model based on Power Set Tree has been introduced. A custom solution designed and built to find the unique combinations for each case in a Case Base. Then use these unique combinations to build the Case Base Index. Finally, a better algorithm has been built to balance the resources consumptions and harness them to serve the purpose of finding the unique combinations for large cases that has more than 38 finding.