1,720,978 research outputs found
Improving the Accuracy of Recommender Systems Through Annealing
No full textCollaborative filtering (CF) is the most popular approach in recommender systems (RS). It makes use of a user item rating matrix and recommends on the basis of preferences and tastes of other users. It faces a number of issues like cold start problem, shilling attack problem and sparse matrix problems. Matrix Factorization (MF) is an efficient approach to get rid of sparse matrix problems. It is a highly reliable and robust technique that helps to predict those ratings to a user for an item that are not yet rated by him. This is done by mapping items and users to a latent space based on a given number of latent features. Minimization in MF is done by either Alternating Least Squares (ALS) method or Stochastic Gradient Descent (SGD) technique.
In this thesis, SGD is used to perform minimization on the matrix factorization function using the concept of singular value decomposition (SVD) to fill in missing entries in the sparse user item rating matrix. Using this base approach, a factor known as learning rate (η) is varied to determine the accuracy and convergence rate of recommender systems. This is done by using simulated annealing, which decrements the value of learning rate in each iteration and provides an optimal solution to minimize error in the system.
In this thesis, five simulated annealing schedules, along with a new proposed annealing schedule have been chosen to discuss the effect of learning rate on the accuracy of a movie recommender system. These annealing schedules are- exponential annealing, inverse scaling logarithmic cooling, linear multiplicative cooling and quadratic multiplicative cooling. Our proposed annealing schedule is named as Square Root Cooling (SRA). The experimental results on Movielens dataset prove that by employing exponential annealing schedule as the learning rate, minimum mean absolute error can be attained for the system at a lower value of learning rate. For higher learning rate values, SRA works the best. Apache Mahout 0.9 is chosen as the platform for the research
Requirement Engineering in Component Selection
No full textM.E. (Software Engineering)In the COTS selection process, it is unavoidable to encounter mismatches between
system requirements and COTS products’ capabilities. These mismatches are the
result of the lacking COTS attributes or their overmatch with the system
requirements.
Many researchers have proposed various techniques to search and select COTS
components. But only few have worked upon the handling of mismatches. These
techniques have their own advantages and disadvantages.
This thesis work proposes a COTS selection approach that aims at addressing COTS
mismatches among COTS attributes and system requirements during the requirement
engineering phase. In this process COTS evaluation, a core activity of COTS
selection, is performed using the Goal Question Metrics (GQM) based Feature
Analysis technique. GQM based Feature Analysis is an evaluation strategy used for
measuring the compliance of the COTS capabilities with desired features using the
formulated metrics. A case-study is used to illustrate this process and to discuss its
feasibility.CSE
Improving Reliance and Efficiency in Scheduling by Development of Hybrid Intelligent Agent
No full textME, CSEDThe thesis includes the research work where it describes the development of a hybrid
intelligent agent in an android application called scheduler cum emailer. The intelligent
agents are the software programs that do some of the intelligent tasks on the human
behalf and acts as its personal assistants to whom user can rely and trust. The intelligent agents are loyal and perform all the tasks to accomplish its designed goal without any mistake. The hybrid intelligent agent developed is the combination of the three types of agents namely: Task agent, triggering agent and goal based agent. The thesis consists of the detailed study of the different types of agents which has been proposed by different researchers and developers. Also, a number of applications and services have been included to give an overview that how the intelligent agents can facilitate and automate the real life human activities by which users can perform his task more easily, timely and efficiently avoiding repetitive and mindless work. The research work includes the application’s salient features and all the intelligent tasks it performs for the user to make
the application more useful and user friendly. The application developed uses the android platform and run only on the android phones. Also the architecture of the application and the proper flow chart of the processing and execution of the algorithm are given which gives the deep insight of how the application is used and works. The application developed is user friendly, easy to use, reliable, efficient, time saving and reduces efforts
Testing Anomalies in Multiple and Multilevel Inheritance
No full textSoftware testing is an important phase of software development process that can be easily
missed by software developers because of their limited time to complete the project.
Thus, Software testing is a widely used and accepted approach for verification and
validation of a software system. One of the major challenges in software testing is the
generation of test cases that satisfy the given competence criterion. Testing in an object
oriented manner significantly increases software reusability, extendibility,
interoperability, and reliability. Object oriented testing has to deal with the new problems
to deal with the new problems introduced by the new features of Object oriented systems
such as inheritance, polymorphism, method overloading, method overriding etc. Object
oriented testing technique for testing software units that has great potential for improving
the quality of testing and to assure the high reliability of software. In this thesis, we will
focus on the features of object oriented systems that create many binding anomalies
during static and dynamic binding. This shows that a value of the variable changes with
the change in the object if there is given a wrong function call with the object of other
class. To detect such anomalies an approach is discussed that will detect static and
dynamic anomalies in multiple and multilevel inheritance
Speech Emotion Recognition Using EEMD, SVM & ANN
No full textME, CSEDEmotion recognition system from speech is one of most advanced topics in the electronic
media. Emotion detection helps the security system to prevent the data from various
attacks at the cyber world. A lot of research work has already been done into this contrast but the problem of accuracy is always there. This work has been done to categorize three emotions namely HAPPY, FEAR AND SAD using the EEMD, SVM and ANN
algorithms. In this work, noise levels are taken so that the emotion can be identified even though if the voice signal is highly noised. The aim of this work is to check the accuracy of the EEMD algorithm with noisy signals in contrast to the emotion detection. We proceed as detecting the noise level and segmenting the signal for the further processing. There are two segments: first part is the training part in which the system is trained to identify the further proceedings. In this part, samples of each voice category are taken and their features are fetched after successful segmentation of the voice file and further on saved into the database. The second part is the testing part in which a voice sample is
taken and all the required properties are fetched and matched with the saved database
values. The closest match comes out as the category of the voice file
An Approach for Improving Accuracy of Prediction Using Ensemble Modeling
No full textM.E. (CSED)In general terms classification can be divided into two steps. First one is learning step which
consists of the predetermined set of classes or concepts. Second step involves testing, in
which data sets are being tested for the verification. And after the system is trained with data
it can be used for further analysis to be done in future so that future events can be predicted in
advance. For different applications we need to apply these models to predict and note the
accuracy given by each model. The main aim of the research here is to make such a system
which has more accuracy as compared to what previous systems are giving. So to implement
this type of system, a hybrid approach is used i.e. ensemble of classifiers. It is necessary that
one does not weigh one model purely. Other models or methods can also give more efficient
results. In doing so, give weightage to each method and combine these methods to reach final
destination that is most informed one. There are a large number of models available which
are used for classifying the data into various class labels. That is also known under various other names, such as multiple classifier systems, committee of lassifiers, or mixture of experts. The basic aim of ensemble based systems is shown to produce favourable results compared to those of single-expert systems for a broad range of applications and under a variety of scenarios. There are various procedures available through which the individual classifiers can be combined. These procedures are called as combination rules. Each rule has its different functionality which will work according to requirement and application where it is applied. So the study is carried on the prediction by applying methods and ensemble that with variable seed values. The experiment is carried out on the k-fold validation to check the consistency of the system
Role of Testing in Phases of SDLC and Quality
No full textM.E.(CSED)Software testing is a technique aimed at evaluating an attribute or capability/usability
of a program or product/system and determining that it meets its quality. Although
crucial to software quality and widely deployed by programmer & testers, software
testing still remains an art, due to limited understanding of the principles of software.
The difficulty in software testing stem from the complexity of software we cannot
completely test a program with moderate complexity. Testing is more than just
debugging. The purpose of testing can be quality assurance, verification and
validation, or reliability estimation. Software testing is a trade off between budget,
time and quality. There is various type of testing, We choose that type of testing
technique which is suitable according to our quality attribute. We are concerned with
the life cycle of finding the debug by which we can easily understand that how the
debugs are found and report it. Quality is the central concern of software engineering.
Testing is the single most widely used approach to ensuring software quality.
This thesis report presents various types of software testing techniques and their
classification. In this report a model “software development life cycle testing model”
is proposed in which we categorise all type of testing techniques related to quality
attribute and to test all phases of SDLC and identify that which type of testing
technique can be applied to which type of SDLC phase.
Software testing is an important technique for assessing the quality of a software
product. In this thesis, various types of software testing technique and various
attributes of software quality are explained. Identifying the types of testing that can be
applied for checking a particular quality attribute is the aim of this thesis report. All
types of testing can not be applied in all phases of software development life cycle.
Which testing types are applicable in which phases of life cycle of software
development is also summarized
Role of OCL and its Supporting Tools in Requirement Specification
No full textM.E. (Software Engineering)During the formal specification phase, the engineer rigorously defines a system using a
modeling language. Modeling languages are fixed grammars which allow users to model
complex structures out of predefined types. This process of formal specification is similar to the process of converting a word problem into algebraic notation. The Object
Constraint Language (OCL) is a notational language for analysis and design of software
systems, which is used in conjunction with the Unified Modeling Language (UML) to
specify the semantics of the building blocks precisely. If the focus of the development
activities is shifted from implementation code to more abstract models then software
developers need a formalism to provide a complete, unambiguous and consistent model
at a very detailed level. So, OCL is currently the only language that can bring this level of detail through specification to UML models. OCL can also be used by other languages,notations, methods and software tools in order to specify restrictions and other expressions of their models. Likewise, OCL is used by the Object Management Group
(OMG) in the definition of other fast spreading industrial standards such as Meta Object Facility (MOF) or XML Metadata Interchange (XMI). The language is very powerful because it can be used together with class and other UML diagrams at different model layers by the specification of OCL constraints at the meta model layer.
In this thesis, work is an attempt to synchronize the language specification and its understanding, straight related to the language improvement in CASE tools, by proposing solutions for incomplete or non deterministic OCL specifications. Also present here is a tool based approach to validating UML models and OCL constraints by implementing a Case Study. The ideas presented in a Case Study have been implemented in the OCLE2.0.4, describe existing principles and stages for generating code from OCL expressions pointing out the drawbacks that cause inefficiencies of the OCL Specifications and its resulting code generated by OCLE2.0.4. The proposed improvement of the code transformation is based on extended Abstract Syntax Trees (AST) with context specific attributes. Finally the feasibility of results presented in this work was shown with the realization of the USE2.4.0 tool. The USE tool has been used to validate the well formedness rules in the UML standard and OCL specifications of a case study. The results provide input for improving future versions.CSE
Designing RIMCOTS Model for Risk Identification and Mitigation for COTS- based Software Development
No full textCOTS-based software development (CBSD) shifts the emphasis from programming
software to composing software systems from off-the-shelf components. Component
based software development is often considered as a low risk development strategy,
which provides a simple and rapid mechanism for developing software systems from
pre-existing and proven components. In reality COTS-based software development
carries significant risks throughout the software development life cycle. These risks
can be due to the nature of COTS components, development process, component
technologies and vendor support. The part of thesis work involves an approach to
design a RIMCOTS model that incorporates risk identification and risk management
mechanism for mitigation of risks in the COTS-based software development.
The purpose of this thesis work is to identify, analyze and explore critical risks
in various stages of COTS-based software development that can affect long-term
viability of software systems.
The Proposed RIMCOTS model includes techniques for evaluating and mitigating
risks associated with deploying COTS components in Component based software
systems and serves as a guide for how to manage multiple COTS software
components in complex COTS-based systems. We validate this proposed RIMCOTS
model by using CURE methodology
Framework for Improvement in Cleanroom Software Engineering
No full textM.E. (Software Engineering)Cleanroom Software Engineering is a theory based team-oriented engineering process
for developing very high quality software under statistical quality control. The
Cleanroom process combines formal methods of object-based box structure
specification and design, function theoretic correctness verification, and statistical
usage testing for reliability certification on to produce software approaching zero
defects. Cleanroom Software Engineering is a management and technical process that
produces high quality software. The focus of Cleanroom Software Engineering
involves moving from traditional, craft-based software development practices to
rigorous, engineering-based practices. Cleanroom Software Engineering yields
software that is correct by mathematically sound design, and software that is certified
by statistically valid testing. In this report detail description of Cleanroom Software
Engineering and Object Oriented Technology is given.
After studying the advantages and flaws of Cleanroom Software Engineering two new
steps have been introduced in the existing Cleanroom Software Engineering. This
report discusses the Collaborative Engineering and Mutation Testing that have been
introduced to improve the Cleanroom Software Engineering. A Comparative study of
Cleanroom Software Engineering and Object Oriented Technology is done and
summarized in a table.CSE
- …
