Abstract: In the present decade, the growth of medical fields with the help of computers opened the new beginning for medical research. In medical research, data mining techniques plays a vital role in predict disease earlier and diagnose. Cancer research is one of the research areas in the medical field. Accurate prediction of different cancer types has great value in providing better treatment and cure the disease all well as possible. The systematic approach is to analyze global gene expression which provides a accurate solution for the identified problem area. In this paper, the input from multiple sources, create a store. Cluster the data with attribute match association rule and followed by classification with the knowledge acquired.

Abstract: In today scenario World Wide Web is flooded with huge amount of information. Finding useful information from the Web is quite challenging task. There are many search engines available in the market that will solve our purpose. However among all, selecting proper search engine with highly effective web crawler is quite necessary. There are many challenges in the design of high performances web crawler like it must be able to download pages at high rate, store it into the database efficiently and also crawl page rapidly. In this paper we present taxonomy of WebCrawler, various challenges and solutions of WebCrawler and various crawling algorithms.

Abstract: Security to any of our personal thing is our most basic need. It is not possible to directly apply that standard Human-computer—interaction approaches. Important usability goal for authentication system is to support users in selecting best passwords. Users often select text-passwords that are easy to remember but they are more open for attackers to guess. Human brain is good in remembering pictures rather than textual characters. So a best alternative is being designed that is Graphical passwords. But Graphical passwords are still immature. Conventional password schemes are also vulnerable to Shoulder-surfing attacks, many shoulder-surfing resistant graphical passwords schemes have been proposed. Next, we have analyze the security and usability of the proposed scheme, and show the resistance of the proposed scheme to shoulder-surfing and different accidental login.

Abstract: In this paper we propose a new decentralized access control scheme for secure data storage in clouds which supports anonymous authentication. The cloud verifies the authenticity of the series without significant knowledge in the user’s identity before storing information. This scheme also has the added feature of access control. In access control scheme only valid users are able to decrypt the stored data/information. This scheme prevents replay attacks also supports creation, modification, and reading information stored in the cloud. These schemes also address user revocation. Moreover, the authentication and access control scheme is decentralized and robust in nature unlike other access control schemes designed for clouds which are centralized. The computation, communication, and storage overheads are comparable to centralized approaches.

Abstract: Global positioning system is globally used for the navigation purpose. GPS is mainly used in military purpose, farming technologies, civil, transportation for example in vehicles and commercial users around the world. This paper describes a practical model for routing and tracking of vehicles in a large area outdoor environment, based on GSM i.e. Global System for Mobile Communication and GPS i.e. Global Positioning System. The supporting device GPS continuously move with the vehicle and records the position of the vehicle. It can be communicate with the help of GSM modern which is installed in both transmitters and receivers, when it is required by the owner. 
GSM model is controlled by 32 bit ARM7 LPC2148. The device will records positions to supervised centre by the SMS or GPRS which can be located in the GOOGLE EARTH and therefore the current position of the vehicle can be known. This research paper describes the working of GPS tracking system and its usefulness in actual world environment. Localisation algorithm, Kalman filter algorithm are some algorithms we will discuss in this paper.

Abstract: The main objective of "Online Application for the Training and Placement" portal is to automate the manual activity involved in Training and Placement programmer. Rich GUI of the system makes the placement activities in the Training and Placement Programmer Department of the college more interactive. The users of the system typically are TPO - Training & Placement Officer, Students, Recruiter, Visitor and Admin.  TPO of the college can manage the details of all students on his console. He can send Email updates to the students about the placement activities.   The System also provides the personal and academic details of a student to the recruiting company on demand. The students should be able to upload the Curriculum Vitae (CV) by logging into the system. Also, they are provided with the feature of editing some fields like Email ID, mobile number and address. Recruiter/Company representatives may also be able to access/search any information uploaded by the Students. The Administrator of portal should be able to add/modify/delete the student/company information which in turn gets stored or modified in a relational database. The system is an internet/intranet based online application that can be accessed throughout the organization and also outside with the proper login credentials are provided. In SDLC many open source tools are used to effectively develop this application. As part of it, MS Visio2010-at Requirement phase, Enterprise Architect and Agro-UML- at Design phase, Eclipse Helios(ER-Master), PREST and CIDE- at Implementation phase, Graph Walker- a Model-based testing (MBT) will be exploited in the Testing phase to automate the testing procedures of various modules.

Abstract: Multipath routing is a very efficient solution which provides more robustness and better load distribution on the network. Nowadays, it is admitted that inter domain peering links represent the main bottleneck of the Internet. Mainly because of lack of coordination between providers, where they use independent and “selfish” routing policies.  For peering carriers, a robust multipath routing coordination framework relies on the multiple-exit discriminator (MED) attribute of Border Gateway Protocol (BGP) as signaling medium. This scheme relies on a game theory modeling, with a non-cooperative potential game by considering both congestions and routing costs. By selecting Pareto-superior Nash equilibrium at each carrier Peering Equilibrium Multi Path (PEMP) coordination policies can be implemented. With the help of this work, the stability of routes can be well improved and congestion can be avoided practically on the peering links. Simulation tool is used to test the performance of the proposed system.

Abstract: Data deduplication is one of important data compression techniques for eliminating duplicate copies of repeating data, and has been widely used in cloud storage to reduce the amount of storage space and save bandwidth. To protect the confidentiality of sensitive data while supporting deduplication, the convergent encryption technique has been proposed to encrypt the data before outsourcing.. In the proposed security model, Security analysis demonstrates that our deduplication systems are secure. The distributed deduplication systems future aim is to reliably store data in the cloud while achieving privacy and consistency.  For achieving data deduplication we propose two types of deduplication methods: block wise data deduplication and file level deduplication with integrity auditing.

Abstract: Android is a java based operating system which runs in the Linux 2.6 kernel. Android applications were developed using java that can be ported to new platform easily thereby fostering huge number of useful mobile applications. The system describes about a SOS application that is being developed and it has been successful implemented. The application whose target users are those sections of the people who falls into situation where instant communication of their network becomes indispensable to be informed certain authorized persons at remote end. The software is developed with the Android App with the automatic notation upon finding important locations using GPS Locator. The application enables the users to find the exact location of the Hospitals, Fire Stations, Police Stations, Blood bank and Ambulance that result from their better applications of technology during emergency periods. This application is used by registering the user details i.e. Name, Contact number, Address then save the details of the concern person. It will navigate the user to desired window using of Android “Smart Phones” and the GPS locator.

Abstract: The numerous applications should work in Large-scale sensor networks domains. The data collected from wireless sensor network are used in making commitment in critical infrastructures. Data's are originated from multiple sources and transmitted through intermediate processing nodes. Those nodes perform the gathering of information. An attacker makes an agreement with those type of networks by introducing additional nodes in the network or compromising the existing nodes. So achieving the high data trustworthiness is crucial for correct decision-making. While evaluating the trustworthiness of sensor data provenance is an important factor. The several challenging requirements for provenance management in sensor networks are low energy and low bandwidth consumption, competent storage and secure transmission. This survey proposes a new lightweight scheme in order to securely transmit provenance with sensor data. The proposed in-packet Bloom filters techniques used to encode provenance with the sensor data. This mechanism initially performs provenance at the base station then perform reconstruction of the data at the base station. In addition to this the provenance scheme functionality used to detect packet drop attacks organized by malicious data forwarding nodes. This survey describes the effectiveness and efficiency of the Light weight secure provenance scheme in detecting packet forgery and packet loss attacks.

Abstract: Text mining refers to the process of procuring high-quality information from text. There are different technologies for text mining. One of the most successful will be mining using the effective patterns. In this paper we have explained one of the most powerful and successful algorithms of association mining named as Apriori algorithm. Lots of algorithms for mining association rules and their variations are proposed on basis of Apriori algorithm, but traditional algorithms are not efficient. Proposed algorithm improves Apriori algorithm by the way of a decrease of pruning operations, which generates item sets by the apriori operation.

Abstract: As it has turned into the standard for cloud suppliers to have numerous datacenters around the sector, crucial requests exist for between datacenter records moves in vast volumes, e.G., relocation of big records. A test emerges on the excellent manner to plot the mass records exchanges at diverse desperation tiers, with a particular stop goal to completely use the reachable among datacenter switch pace. The software program described Networking (SDN) worldview has risen as of late which decouples the control aircraft from the statistics methods, empowering ability worldwide advancement of statistics directing in a system. This paper plans to define an element, highly productive mass records circulate management in a geo-circulated datacenter framework, and dressmaker its configuration and association calculations almost internal of a SDN engineering. We exhibit statistics alternate requests as deferral tolerant relocation needs with diverse completing due dates. Because of the adaptability gave through SDN, we empower changing, perfect directing of unmistakable pieces inside of each mass records exchange (in preference to concerning each change as an endless flow), which can be in short placed away at middle of the road datacenters to relieve transmission ability conflict with more earnest exchanges. A really perfect piece directing streamlining version is deliberate to fathom for the exceptional lump alternate plans after some time. To determine the appropriate calendars in an internet way, 3 calculations are pointed out, to be precise a switch velocity saving calculation, a regularly altering calculation, and a future-demand friendly calculation, focusing at diverse ranges of optimality and flexibility. We manufacture a SDN framework contemplating the Beacon stage and OpenFlow APIs, and painstakingly design our mass information move calculations within the framework. Broad certifiable tests are conveyed out to examine the three calculate the present day writing, as some distance as steerage optimality, computational postponement and overhead.

Abstract: Mobile Voting System (MVS) is a very secure, efficient and easy way to casting of vote. Mobile Voting System creating Online Voting System. All user will enter their details in verification form, server will check whether they are valid user or not. If it is valid, user server provide user Id and Password. By using Application Program Interface (API) and generating One Time Password  number administrator will send to the user by using One Time Password number Encryption and Decryption and also visual cryptography. Security providing double time encryption and decryption using visual cryptography. If the hacker tried to hack it will be displayed image or video. System provide a new Electronic Voting (E-Voting) system which full fills the security requirements of voting process. In this paper total three steps are required: Online Registration of the  vote casting  system for voters and displaying the result. There is  a need for the usage of internet  for voting purpose. It is required at the time of online registration only. All process will be done through Short Message Service  messaging with requirement of internet connection.

Abstract: Android applications are open source and can be developed by anybody, testing is not done. The objective of this paper is to remove the unused/redundant permissions in the android applications by breaking it and extracting the permissions to prevent the permission gap. High level Permission Checking Framework on Android Applications that were previously uploaded by breaking the .apk files to analyze in code level by decompiling it in a efficient way. This is a compositional analysis for Android inter app vulnerabilities. Both developer and user can test and modify manifest file of an application.

Abstract: The paper is intended to introduce some well known sorting techniques using array and C++. The paper briefly defines and describes the different sorting techniques available and well known to us giving suitable example.

Abstract: The Internet of Things is an emerging phenomenon of today having great significance, technology, and global economy point of view. The extensive implementation of Internet of Things is ready to transform our life and lifestyle. Internet of Things offers huge potential due to unprecedented growth of smart devices, availability of low cost technologies, comprehensive connectivity and massive volumes of data collecting and processing ability in real time. Internet of Things is the next generation of the Internet connecting things and devices like sensors and cameras, results in generation of huge data that can infer useful insights for decision making. A world of interconnected smart devices is here in the form of Internet of Things. Internet of Things raises many issues and challenges like risk of surveillance, privacy and security violations. The present paper discusses the concept, definitions and application areas to gives an overview of the emerging technology Internet of Things.

Abstract: Four Square is one of Geo spatial analysis in which lots of communities interrelate with their surrounding environment through their friends and recommendations. We must have to take care of our data and Location. But this is not possible every time sometimes because of busy time we can’t take care of our data without sufficient location protection. However, these application can be easily misused. So, we proposed enhanced location privacy policies in Geo spatial networks. A technique that provides location secrecy without adding complexity into query results. Idea here is to secure user specific data and coordinate conversation to all location data shared with the database server. To provide privacy in server using Location index mapping and distance preserving coordinate transformation. The associates of a user share this user’s secret key. This allows all spatial queries to be evaluated correctly by the server but privacy mechanisms guarantee that servers are unable to infer the actual location data from the transformed data and easily through the data access.

Abstract: FPGAs have achieved quick acceptance, spread and growth over the past years because they can be applied to a variety of applications. Some of these applications includes: random logic, bioinformatics, video and image processing, device controllers, communication encoding, modulation, and filtering, limited size systems with RAM blocks, and many more. For example, for video and image processing application it is very difficult and time consuming to use traditional HDL languages, so it’s obligatory to search for other efficient, synthesis tools to implement your design. The question is what is the best comparable language or tool to implement desired application. Also this research is very helpful for language developers to know strength points, weakness points, ease of use and efficiency of each tool or language. This research faced many challenges one of them is that there is no complete reference of all FPGA languages and tools, also available references and guides are few and almost not good. Searching for a simple example to learn some of these tools or languages would be a time consuming. This paper represents a review study or guide of almost all PLD's languages, interpreters and tools that can be used for programming, simulating and synthesizing PLD's for analog, digital & mixed signals and systems supported with simple examples. In addition, their features and applications will be discussed. In Addition to the aforementioned summary, the paper presents a new classification for FPGA languages and Tools as well. Interpreters like PERL and MYHHDL (python based) also described here. At the end one can summarize all what’s needed to know about FPGA languages, Tools, Compilers, and Interpreters in one research paper.

Abstract: Cloud computing is the way of sharing the computer resource over the Internet and hence it is termed as the Internet based computing. But the problem with cloud is that many user store their data in cloud it has some security issues like collusion attack, loss of data. In existing system, they used the concept of user revocation to secure the data and to avoid collusion of the data using ECC algorithm which uses elliptical equations and complex calculations. In proposed system, we are using AGKA(Asymmetric Group Key Agreement) algorithm for encryption and decryption. For efficient user revocation we are using verifier local group signature method. And to maintain the integrity of the data TPA (Third Party Auditor) is used for checking the data. Our scheme provide some properties, such as confidently, efficiency, countability and traceability of secure group user revocation. Finally our scheme achieve the integrity and security.

Abstract: In modern days, size of datasets becomes larger and larger. The conventional tools and techniques cannot able to handle those large data. In order to extract value from massive amount of data, frequent itemset mining is an important tool which could give better insights for the business environment. Frequent itemset mining is the most popular technique for data mining, but main memory was not capable to load these large datasets. The main goal of frequent itemset mining is to extract frequent itemset MapReduce is used to overcome this limitation by using parallel processing of Big Data which helps to handle problem of large datasets. The proposed ClustBigFIM works based on MapReduce framework for mining large datasets ClustBigFIM is modified BigFIM algorithm providing extensibility and speed in order to reveal useful information from large datasets. ClustBigFIM is the hybrid approach which is the combination of apriori and eclat algorithm. The proposed algorithm uses k-means clustering algorithm for clustering, apriori and eclat to mine frequent itemsets. ClustBigFIM algorithm is used to make effective business decisions in competitive business environment by using the patterns obtained from the mining process.

Abstract: The Android is a growing technology which was started to fulfils various needs. Friend Book is a mobile Text messaging Application and Cross-platform Application designed to let cell phone users send text messages without having to pay Short Message Service amounts to their carrier. The message are sent via Internet. Is a real time Chatting application also sometimes as real-time text transmission over the Internet. First, the User register in to Friend Book, and start synchronous chat. Synchronous chat is text only web based synchronous forum that enables multiple user to be On line and typing their comments to each other. Both the persons may communicate through Friend Book App. The message may be text,video,audio or Portable Document Format file. The users of the Friend Book may join together and form the group. Google Play Service is used to maximize the privacy, manage the accounts, synchronize the contacts, Identify the location and update the location with map. It provides accurate location to user.

Abstract: Artificial Neural Networks (ANNs) are non-linear data driven self adaptive approach as opposed to the traditional model based methods. They are powerful tools for modeling, especially when the underlying data relationship is never known. In recent years, Neural Networks have become increasingly popular in finance for tasks such as pattern recognition, classification and time series forecasting. The motivation of this study is to use recurrent dynamic neural networks in enhancing the ability of GARCH models in forecasting the return volatility. A monthly data of Naira/Dollar exchange rate (January, 1991 to January, 2016) and inflation rate (January, 1970 to December, 2015) obtained from central bank of Nigeria’s website www.cbn.gov.ng were used in this research.  Hybrid models were developed based on the preferred GARCH models and Recurrent Neural Networks to forecast the volatility of these variables. The estimates of the volatilities of the preferred GARCH models were used as inputs and monthly standard deviation as the target of the proposed models. The input to the hybrid I model also included values of other related variables. Simulated series were used as complementary inputs to hybrid model II. The computational results show that GJR-GARCH (1, 2) fits best the Naira/Dollar exchange rate return volatility and EGARCH (2,3) fits best inflation rate returns volatility. The forecasts obtained from these GARCH models and the hybrid models were compared in terms of the standard deviation as a measure of the actual volatility and the results demonstrate that the second hybrid model provides better volatility forecasts.

Abstract: "Automated Blood Bank system" brings voluntary blood donors and those in need of blood on to a common platform. This project servicing the persons who seek donors who are willing to donate blood. “Direct contact between the donor and recipient” is the main concept of this project. Direct call routing is the technique used for providing the communication between the donor and the recipient. A blood bank database is created by collection of detail through web interface. The data collected will be maintained in a central server. By using eligible donor finding algorithm the most eligible donor is found out and the call will be routed. From the server  the  call  from  the  required  person  is  routed  to  the eligible  donor's  number. Willingness of the donor, Location, Last donation date and blood group are the major four components considered in this project.

Abstract: Haze is otherwise called as fog or snow. Outdoor images captured during inclement weather conditions exhibit visibility degradation (ex: presence of haze, sandstorms, camera on motion, blur streetlights and headlights).While removing the haze, the visibility of the scene get increased and with the correct color shift caused by the atmospheric light. In general the haze free image is more visually pleasuring the scene. Haze removal can provide depth information and benefit many vision algorithms and advanced image editing. The bad haze image can be transformed to good use. In that we propose a new CLAHE de-haze algorithm for removing the fog or de-hazing the image by removing the dark channel prior & atmospheric light. To enhance the de-haze image by smoothing the image using CLAHE based histogram equalization method is applied to it. Good and enhanced image with better scene radiance is obtained as the final output.

Abstract: The DDOS Attack in a Client Server Environment would Collapse the Entire System, but as far as Cloud is concern it is not that Effective but still it will try to Disturb the Regular Activity of the System. The proposed system, User’s Request could also increase the CPU Load of the Cloud Server. Filters the Request based on the behavior and forwards to the corresponding Servers through Cloud Server. Every Server would have allocated Certain Space in Cloud Server. Our system monitors the Activity of the Users to Avoid DDOS Attacks. We Deploy once data owner upload their data along with their keyword data stored in separate server. We implement seven types of attacks. 1. Continuous & same request from single user in a point of time. 2. Different query from the same user within a period of time. 3. Different queries from different users but from same IP. 4. Request of huge sized file beyond the permitted. 5. Wrong user name & password for more than 4 times. 6. IP address of the registered user. 7. CPU utilization of the user’s request. Based on these pattern user behavior is monitored DDOS attack is avoided in cloud. Mango lab is also implemented.

Abstract: Cybercrime is spreading all over the globe very rapidly. Cybercriminals are increasingly inventing innovative ways of committing Cybercrime. The worldwide trend of certain types of Cybercrimes is much ahead other due extensive amount money they generate. The Cyberdefence need to be more perfect attempting to eliminate all sorts of vulnerabilities by adopting a risk based approach. Cyberattacks are real; they potentially impacts and so nobody can ignore increasing threat of Cyberattacks. The current research paper discusses some of the leading and severe Cybercrimes.

Abstract: There is a rapid growth in E-Commerce market in recent time throughout the world. Also with the ever increasing popularity of online shopping, Debit or Credit card fraud and personal information security are major concerns for customers, merchants and banks specifically in the case of CNP (Card Not Present). This paper presents a new approach. Only limited information is provided which is necessary for fund transfer during online shopping thereby safeguarding customer data and increasing customer confidence and preventing identity theft. The method uses combined application of visual cryptography and steganography for this purpose.

Abstract: Disruption Tolerant Network is a network architecture that reduces intermittent communication issues by addressing the problems in networks that cause lack of continuous connectivity. It played a major role in military environment. In military environments connection of wireless devices carried by soldiers may disconnected due to some environmental factors. Disruption Tolerant Network has been introduced to provide a successful solution for that problem. Message Authentication Code (MAC) technique has been used to maintaining the integrity of the message from source to destination. For each round of Message Authentication Code (MAC) generation different key has been generated by Key Distribution Centre (KDC). Here, the message has been encrypted and decrypted using Advanced Encryption Standard (AES) algorithm. Advanced Encryption Standard is a symmetric block cipher that will using the same key for both encryption and decryption at the sender and receiver side.

Abstract: This paper is intended to describe and discuss the major developments of a website design and internet that makes a website much more attractive, useful, an effective business tool and much more powerful than ever before. The paper is basically designed and prepared for the aspiring web designers and developers with a need to understand the HTML as well as website design in enough detail.

Abstract: Computer may be the target of the offense; attacks on network confidentiality, integrity and / or availability i.e. unauthorized access to and illicit tampering with systems, programs or data or computers may be in supplementary role in the commission of a traditional offense. Cybercrimes range from economic offenses fraud, theft, espionage, sabotage and extortion, product piracy, etc. to infringements on privacy, propagation of illegal and harmful content, facilitation of prostitution and other moral offenses, and organized crime. At its most severe, Cybercrime borders on terrorism, encompassing attacks against human life and against national security establishments, critical infrastructure, and other vital veins of society.

Abstract: Lack of information about irrigation potential has caused crop failures on irrigated farms in Kenya due to unexpected climate change. The information model through annual reports in form of table and text as provided by agricultural and meteorological experts is clearly not understood by the farmers. Therefore, this paper aimed to visualize irrigation potential by modeling the rainfall and runoff using Microsoft excel. Using Microsoft 2007 in a normal computer, daily rainfall was recorded into one sheet, calculated into another sheet as runoff and visualized in the last sheet as a map of irrigation potential. One sheet was added to record actual runoff harvested in a reservoir and another to distribute the runoff onto irrigation areas. A raster map was visualized as a result of modeling the entered data implying that ease of information could be achieved almost at no extra cost to what is usually done.

Abstract: One of most challenging problem is the replica attack in static wireless sensor network. Also every sensor nodes are physically captured. These nodes are reprogramming and replicated in large number of replicas. Which may dynamically occupy the network Thus far different ways to detect the replicas? Most of the sensor nodes required high costs hardware like as: “Global Positioning System”. In general, Sensor nodes are low price as compared to GPS hardware. On this paper, we proposed "Low Priced and Energy-Efficient Detection of Replicas in Static Wireless Sensor Network”. On this proposed solution not required any internal hardware such as: GPS. Good performances as compared to exiting system. We show that the proposed solution saves the lot of energy than exiting system.

Abstract: Business intelligence (BI) and healthcare analytics are emerging technologies that provide analytical capability to help healthcare industry improve service quality, reduce cost, and manage risks. However, such component on analytical healthcare data processing is largely missed from current healthcare information technology or health informatics curricula. The paper refers to the use of services like SSIS, SSAS, SSRS to provide integration, analysis and reporting of humongous data, respectively. The project conducts a analysis on how business intelligence can be incorporated in the field of healthcare informatics for betterment, enhancement in the field .A general framework and exemplar implementation strategy is presented.

Abstract: In India there are different agriculture crops production and those crops depends on the various kind of factors such as biology, economy and also the geographical factors. And this several factors have the huge different impact on crops, which can be quantified using appropriate statistical methodologies. Applying such methodologies and techniques on historical yield of different crops, it is possible to obtain information or knowledge which can be helpful to farmers and government organizations for making better decisions and for make better policies which help to increased production. In this paper, our focus is on application of data mining techniques which is use to extract knowledge from the agricultural data to estimate better crop yield for major crops in major districts of India.

Abstract: Class imbalance is one of the critical problems in classification. Because when classifiers classifies the records in imbalanced datasets, the  classified datasets provides best classification accuracy for class(es) which has more number of examples and poor classification accuracy for the class(es). The poor classes have less number of examples. Therefore solving the class imbalance before applying the classification algorithm or at the time of applying the classification algorithm is very important, if good classification accuracy for minority class (es) (class with less number of examples) is needed. To identify the better approach to solve class imbalance problem, in this project the algorithm is compared with performance of four ensemble methods (Adaboost, bagging, random subspace & rotation forest). To evaluate the performance of the algorithm the datasets with severe, less severe imbalance ratios are used.

Abstract: Web applications use recommendation techniques that are based on users' preferences for items to recommend interesting items to the active user. Users' preferences can be their activities on these items such as: rate, view, etc. Collaborative filtering (CF) is the most successfully used recommendation system recently; it is based on the similarities between users and the similarities between items. However, users' opinions and items' popularities vary with time. These variations decrease the recommendation accuracy. Many researchers study ways of using Morkov model in recommendation systems; however, they do not consider the time factor in their techniques. In this paper, we propose a new recommendation system that is based on Morkov model, and it considers the time factor. We compare our new technique with the conventional CF recommendation system for the evaluation. We conduct the experiments using MovieLens dataset. The evaluation is done by using precision-recall, and mean absolute error. The result illustrates that our new recommendation system outperforms the conventional CF recommendation system.

Abstract: Web applications use recommendation techniques, which are based on users' preferences for items, to recommend interesting items to the active user. Many researchers study ways of using Morkov model in recommendation systems; however, they do not consider the time factor in their techniques, while users’ opinions and items popularities vary with time. In our previous paper, we propose a new recommendation system that is based on Morkov model, Markov Chain Recommendation System (MCRS), and it considers the time factor. In this paper, we enhance MCRS using the general weights of items, and extra enhancement is done using period weights of items. The new enhanced techniques are compared with MCRS for the evaluation. The experiments are conducted using MovieLens dataset for the evaluations that use the precision and recall, and the recommendation accuracy to compare between MCRS and its enhancements. The result illustrates that the enhancements techniques outperform MCRS.