Artificial Intelligence is the key technology in new applications, ranging from banking systems that detect attempted credit card fraud, to telephone systems that understand speech, to software systems that notice when you're having problems and offer appropriate advice. These technologies would not exist today without the sustained federal support of fundamental artificial intelligence research. At present, knowledge-based system technology requires the means to provide efficient and robust knowledge bases, while database system technology lacks knowledge representation and reasoning capabilities. The time has come to face the complex research problems that must be solved before we can design and implement real, large scale software systems that depend on knowledge-based processing. The essential of technology use operational level for knowledge-based system applications. In this article the author discussed artificial intelligence and knowledge based systems technologies and their applications.
The Real life data is often dirty. To clean the data, efficient algorithms for detecting errors have to be in place. Errors in the data are typically detected as violations of constraints (Data quality rules), such as Functional Dependencies (FDs), Denial Constraints, and Conditional Functional Dependencies (CFDs). When the data is in a centralized database, it is known that two SQL queries be adequate to detect its violations of a set of CFDs. It is increasingly common to find data partitioned vertically or horizontally, and distributed across different sites. This is highlighted by the recent interests in SaaS of Cloud computing, Map Reduce, and columnar DBMS. In the distributed settings, however, it is much harder to detect errors in the data. To find violations in both settings, it is necessary to ship data from one site to another. It is NP-complete, to find violations of CFDs, with minimum data shipment, in a distributed relation that is partitioned either horizontally or vertically. So, the proposed work introduces such incremental algorithms for vertically and horizontally partitioned data, and show that the algorithms are absolute. Further, propose an optimization technique for the incremental algorithm over vertical partitions to reduce data shipment for error detection.
Mobile nodes in military environments like a battlefield or a hostile region square measure possible to suffer from intermittent network property and frequent partitions. Disruption-tolerant network (DTN) technologies are getting successful solutions that permit wireless devices carried by troopers to speak with one another and access the counseling or command reliably by exploiting auxiliary storage nodes. a number of the most difficult problems during this state of affairs square measure the social control of authorization policies and also the policies update for secure knowledge retrieval. Cipher text-policy attribute-based coding (CP-ABE) is a promising science answer to the access management problems. During this paper, we have a tendency to propose a secure data retrieval theme victimization CP-ABE for suburbanized DTNs where multiple key authorities manage their attributes severally. We demonstrate the way to apply the planned mechanism to securely and with efficiency manage the confidential knowledge distributed in the disruption-tolerant military network.
We study the matter of on-line multi-task learning for finding multiple connected classification tasks in parallel, aiming at classifying each sequence of knowledge received by every task accurately and expeditiously. One sensible example of on-line multi-task learning is the micro-blog sentiment detection on a gaggle of users that classifies micro-blog posts generated by every user into emotional or non-emotional classes. Initial of all, to satisfy the vital requirements of on-line applications, an extremely economical and scalable classification resolution which will create immediate predictions with low learning value is required. This demand leaves standard batch learning algorithms out of thought. Second, classical classification strategies, be it batch or on-line, typically encounter a quandary once applied to a gaggle of tasks, i.e., on one hand, a single classification model trained on the complete assortment of knowledge from all tasks could fail to capture characteristics of individual task; on the other hand, a model trained severally on individual tasks could suffer from insufficient coaching information. To beat these challenges, during this paper, we tend to propose a cooperative on-line multi-task learning methodology that learns a world model over the complete data of all tasks. We also evaluate it on 3 real-life problems—spam email filtering, bioinformatics information classification, and micro-blog sentiment detection. Experimental results show that our methodology is effective and scalable at the web classification of multiple connected tasks.
Software requirement specification is a very important early phase in the software development life cycle. There must bemaintaining consistency between source code and document of a system to produce a quality product. Traceability can be used for showing that source code of a system is consistent with the requirements. Developers do not focus on the traceability links during maintenance and evolution. Later on, recovering these traceability links is difficult.Previous studies focused on recovery of traceability links Information Retrieval (IR) techniques. But IR techniques produce less accurate results in finding traceability links. In the proposed work must be improve the accuracy (precision and recall) through trustrace approach: trust based traceability links(reconstruction of traceability links) and comparing the results of IR techniques and Trustrace.
Authentication is the mechanism in which the system check the identity of the user who access it. Traditional alphanumeric passwords are used. Due to the limitation of human memory, most users try to choose short-term or simple passwords which are easy to remember. Surveys show that frequent passwords are personal names of family members, birth date, or common words. In most cases, these passwords are easy to guess and vulnerable to dictionary attack.In an attempt to create more memorable passwords, graphical password systems have been devised. In these systems, authentication is based on clicking on images rather than typing alphanumeric strings. This paper presents a survey on some graphical password authentication systems
