Analysis and Prediction of Application Usage (APAU)

Team Member Are :
Asha Shenoy K
Macwin Alwyn Lewis
Vishal Monteiro
Namrata

Project Guide: Mr. Shreenath Acharya Abstract: Predictive analytics analyze current and historical facts to make predictions about future events using techniques like machine learning and data mining. Predictive models look for certain relationships and patterns that usually lead to certain behaviour; by determining the explanatory variables, you can predict outcomes in the dependent variables. This concept when applied to real time applications helps users reduce effort and also save time. Smart phones have evolved from communication devices to indispensable accessories with access to real-time content. The ability to foresee what mobile users want to do has many applications, one of which is providing relevant recommendations. A typical smart phone user uses many applications repeatedly like making calls, messaging, checking mails, etc. It is observed that every user exhibits certain patterns in application usage on a Smartphone. The prime objective of our project is to identify these patterns and investigate the prediction of human behaviour while using applications. It deals with extracting generic (i.e., user independent) behavioural patterns and studying how generic behaviour models can improve the predictive performance of personalized models. In this system we make use of the usage statistics of a particular user and predict which application the user will need to use next, thus providing trouble free application switching at a user level. Salient features: • APAU helps users in quick app switching and multi-tasking by providing an easy interface to do so. • APAU is an app recommendation system that takes into account the application usage patterns of an Android phone user. Based on usage statistics, the next application that the user is most likely to use will be predicted. • When APAU is first installed on an Android device, it functions with an initial database that contains generic patterns. This data is used to generate recommendations originally. As time progresses, APAU learns the usage patterns of the user, modifies its database and provides relevant recommendations to the user. • The recommendations are displayed via notifications. This makes the recommendations easily accessible because notifications can be accessed by the user from any application. • User response (acceptance of recommendation) is also recorded to make preference updates in the database so as to recommend the right applications on the next instance.

Multilayered Security Approach for Data Transmission

Team Member Are :
Anoosha Shetty
Averil Nathalie Noronha
Shwetha Nayak

Project Guide: Mr.Bhargav Bhatkalkar, Assistant Professor, ISE Department Abstract: With the huge growth of computer networks and advancement in technology, a huge amount of information is being exchanged between two organizations or individuals through a network which may be insecure. A large part of this information is confidential or private which can be modified or eaves dropped or sensed by intruders during the transmission. The purpose of the proposed project is to provide a full proof security to information being communicated over a network by providing multilayered security by combining Steganography and Cryptography techniques that provide a strong background to the information’s security. Main features: • Both Cryptography and Steganography when used in isolation are vulnerable. The project combines the best features of both the techniques to make data transmission safe and secure. • While using steganography to hide the secret message in it, the Least Significant Bit (LSB) of the cover image is modified. Hence the change in pixel values is not directly evident. • Use of safer encryption and decryption algorithm. • First layer of security is provided by the use of Diffie Hellman key exchange algorithm which generates keys for cryptography and are shared between two peers. • Second layer of security is provided by encrypting the secret message to be sent using AES algorithm. • Third layer of security is provided by using a unique way of extracting pixels from an image to hide the secret data. • Third layer of security is provided by encrypting the stego-image again using AES algorithm. • No degradation in quality of the image and extracted message is same as that which is sent. • Confidentiality and Integrity of the secret message is protected.

Search Engine Aggregator (Sea 1.0)

Team Member Are :
Karthik Mohan
Maithri K
Neha Rakshitha
Tripthi Kuckian

Project Guide: Ms. Gayana M.N., Assistant Professor, IS&E Department Abstract: Search Engine Aggregator aka SEA is a system to shift through the search results of all major search engines available online on the internet aggregate and display it in a search software based on a computer desktop. This particular type of software has never been implemented before, so SEA becomes a pioneer in the clan of software’s to be. SEA at its very best will allow the user to compare between multiple search engines and their efficiency, the user will be given a set of results based on relevance and SEA at its initial form will prioritize the search results based on popular content. The system use well-known programming language JAVA along with jdk and ide environments. SEA is a flexible software. More number of search engines can be added further in future which provides variety of choices for the end users. Further improvements to SEA could include concept of content based searching, live trend feeds and customized profile based searching. This software can be implemented any platform, hence it is a portable software. Features: • SEA is a replacement to the manually driven search on every existing searchable website. • SEA will conduct aggregated search on a set of websites at press of SEA button. • For the users who are not familiar with any other search engine than google, SEA would introduce a whole new part of internet to them • SEA provides priority searches based on 2 levels of priority which could refine the data collected over the internet. • The user level priority is given based on the profile selection and service selection which accomodates the user’s immediate requirements. • Engine level priority is set of hardcoded search urls provided for each user selection. • The modern searches always require an element of meta search, SEA provides this facility by letting the user query a set of meta search engines on the press of a button.

Traffic Sign Recognition Using Color Filtering And Area Differentiation

Team Member Are :
Amitha Rose Thomas
Anusha
Caroline D’Souza
Esha Maria Mendonca

Project Guide: Ms Vijetha U, Assistant Professor, ISE Department Abstract In this project a comprehensive approach to traffic sign detection and recognition is proposed. An RGB roadside image is acquired. Color filtering is performed to detect the boundary of traffic sign in binary mode. At the feature extraction stage, the binary traffic sign region is cropped and image is resized to appropriate pixels. Segmentation of extracted image is performed in order to perform area differentiation. Finally, the value obtained using area differentiation is compared with values for traffic signs stored in database, which is used to identify the traffic sign. Experimental results show that our system can give a high recognition rate for general warning signs of traffic sign. Main features • High recognition rate for general warning signs • Result can be represented visually as well with audio • Traffic signs in almost all formats can be recognized(GIF,JPEG,BMP,PNG etc) • If there are impurities in the input traffic signs, it will discard those impurities and recognize the sign • Further , this can be applied in auto piloting systems as well as in vehicles to avoid road accidents

Automated Remote Software Installation/Uninstallation on Windows Machines (ARSIU)

Team Member Are :
Akila C
Ashritha B J
Soujanya M J
Vidyashree Bhat

Project Guide: Ms Remul Pinto, Assistant Professor, ISE Department Abstract: ARSIU is a software system developed for system administrators where server and client machines are installed with Windows Operating system. It avoids the need for a system administrator to manage each and every client machines connected to the server manually. It includes installing, uninstalling, updating, copying the required software’s as well as shutdown and restartof the client machines. It provides a graphical user interface for the administrator at the server. The server and the client machines must be connected via network. ARSIU is installed on server machine and the appropriate client program runs on the client machines. When the administrator want to perform any operation from server he/she must select the clients on which the operation need to performed and follow the appropriate steps required for the operation. Features of ARSIU: • INSTALL: If Administrator selects the Install operation he/she will be provided with a window to select the software to be installed. • UNINSTALL: If Administrator selects the uninstall operation he/she will be provided with a window to select the software to be uninstalled. • COPY: If Administrator selects the Copy operation he/she is given the choice to Copy the particular software. • UPDATE: If Administrator selects the Update operation he/she is given the choice to update the particular software. • SHUTDOWN: If Administrator selects the Shutdown operation he/she is given the choice to shut down the client systems. • RESTART: If Administrator selects the Restart operation he/she is given the choice to restart the client systems.