International Journal of Engineering Trends and Technology (IJETT) – Volume 22 Number 1- April 2015
Mining Specifications for Error Detection to Improve
Code Quality
Boddepalli VysaGeetha1, T.Ravi Kumar2
1
1,2
Final M.Tech Student,2 Sr.Assistant Professor
Dept of Computer Science and Engineering, Aditya Institute of Technology And Management,Tekkali, Srikakulam.
Abstract:
Automatic specifications from the code
blocks is always an interesting research issue in the
field of software engineering and data mining,
specifications should be accurate and reliable, but in
this specification mining, most of the research works
are false positives.In this paper we are proposing an
efficient automatic specification mining,our first
experiment provides empirical evidence that our
quality metrics are distinct. Our second experiment
presents empirical evidence that our quality metrics
improve an existing technique for automatic
specification mining.
I. INTRODUCTION
Traditional testing process is time consuming
and expensive while handling of large projects. The
traditional approaches are fault prediction works with
the basic metrics like Lines of code(LOC),Number of
errors found, Number of errors found respect to the
module, These parameters are not sufficient to
measure the fault prediction and cost effectiveness.
software faults may be design faults which are
deterministic in nature and are identified easily and
other type of software faults is classified as being
temporary internal faults that are transient and are
difficult to be detected through testing, It is difficult
to analyze the fault prediction by simply measuring
the software metrics of the project, we require a
classification tool for the analysis of the predicted
results.
We can improve our traditional work by
enhancing
the
classification
approach,
In
classification based approach, analysis fails when
testing sample of data not available in training dataset
or new data sample and classification fails when data
is inconsistent or not available for specific attributes
or metrics . By improving these two features we can
enhance the performance of the current proposed
system.
The chances of occurrence of error is more in
frequently or recently modified source code is shown
by the authors[4][5],it is may be due to inter
connection of code block, fixing of one block may
disturbs another block in the source code. So we need
ISSN: 2231-5381
maintain log report which stores number of versions,
modifications or rectifications taken place and that
differentiate the older version and later versions for
each line of code, such metrics can improves chances
of correctness.
Author Rank:
We hypothesize that the author of a piece of
code influences its quality. A senior developer who is
very familiar with the project and has performed
many edits may be more familiar with the project’s
invariants than a less experienced developer. Source
control histories track the author of each change. The
rankof an author is defined as the percentage of all
changes to the repository ever committed by that
author. We record the rank of the last author to touch
each line of code. While author rank may be led
astray by certain methodologies (e.g., some projects
may have a small set of committers that commit on
behalf of more than one author [18]; others may
assign more difficult and thus error-prone tasks to
more senior developers), we note that it may be
automatically collected from version control histories
and is a proxy for expertise, which is otherwise
challenging to approximate automatically.
II. RELATED WORK
Number
of
traditional
programming
languages is not formally determined. Formal
specifications are problematic for the people who
work on it, to investigate and alter the specification
and In like manner analysts have created methods to
consequently induce particulars from system source
code or execution as follows. These techniques
transfers’ solutions as limited state machines that
shows other system practices
Techniques like sql injections are error prone
for the software because it is vulnerable while
appending the strings or it can be followed by the
other queries. To resolve the issue like these we use
parameter approach for secure parameter passing
between the functional calls and we allude to such
properties as determinations for the rest of this article.
Such determinations can be spoken to as a limited
http://www.ijettjournal.org
Page 10
International Journal of Engineering Trends and Technology (IJETT) – Volume 22 Number 1- April 2015
state machine that encodes legitimate successions of
occasions. Data base cannot check the originality of
the query, because it checks whether it is a string or
not, nothing more information over it
A project execution holds fast to a given
particular if furthermore, just on the off chance that it
ends with the comparing state machine in a tolerant
state (where the machine begins in its begin state at
project instatement). Something else, the project
damages the particular and contains a mistake. This
kind of incomplete rightness detail is particular from,
and reciprocal to, full formal conduct determinations.
They can be utilized to depict numerous vital
accuracy properties, including asset administration,
locking, security, abnormal state invariants, memory
security, and then some specific properties for
example the right treatment of set uid or nonconcurrent I/O demand parcels. Such particulars are
utilized by all current deformity finding apparatuses.
Moreover, formal particulars are instrumental in
project
advancement,testing,refactoring,
documentation, also, repair.
III. PROPOSED WORK
We are proposing an empirical model for measuring
the code quality, by extracting the information from
the code through engineering process. Our technique
identifies which input is most indicative of correct
program behavior, which allows off-the-shelf
techniques to learnthe same number of specifications
using only 45% of their original input.Th main
advantage of the proposed system isSpecification
Mining, Code Readability, Path Density and QualityBased Specification Mining.
Path Density
We can define path density
by computing the
number of nodes and a number of edges between the
connected paths and cyclomatic complexity can be
computed with volume and density and define the
path density as metric and number of traces to
enumerate and it can be computed along with all
object oriented metrics.
Quality-Based Specification Mining
Our main experiment measures the efficiency of
ournew specification miner.A leave-one-out analysis
showsthe including the CK metrics in the model
raises boththe true and false positive rate. As our goal
is usefulspecifications with few false positives, we
omit features,even those that are predictive for true
positives thatincrease the false positive rate
substantially.
Our miner takes as input:
1) The program source code P. The variable _ ranges
over source code locations. The variable l represents a
set of locations.
2) A set of quality metrics M1 . . . Mq. Quality
metrics may map either individual locations _ to
measurements, with Mi ∈ R (e.g., code churn) or
entire traces to measurements, where Mi(l) ∈ R (e.g.,
path feasibility).
3) A set of important events Σ, generally taken to be
all of the function calls in P. We use the variables a,
b, etc., to range over Σ.
Our miner produces as output a set of candidate
Specifications C = { a,b | a should be followed by b}.
We manually evaluate candidate specification validity
IV. CONCLUSION
Specification Mining
In Specification mining, specifications can be
constructed doe the actual program or source code
based on the existing sourcecodes, it can be in terms
of function calls and other software metricsor any
other events. It can be static or dynamic approach of
gathering the software metrics from the available or
existing source code blocks.
We are concluding our current research with
specifications of the system, earlier manual system is
difficult to identify the section automatically by
mining it and gives more false positive chances. Our
mechanism computes cyclomatic complexity metrics,
lines of code and object oriented metrics for metrics
calculation required for automatic specifications of
mining and gives efficient and reliable results.
Code Readability
REFERENCES
Reliability is the important factor rate of
success or failure based on the metrics used to
measure the software code quality, these metrics may
differ from the quality metric approach, operators,
operands, comments, number of classes, inherited
class, reference objects and many other metrics.
Reliability factor of probability lies between 0 to 1.
[1] Assessing the Cost Effectiveness of
Fault
Prediction in Acceptance Testing AkitoMonden,
akuma Hayashi, Shoji Shinoda, KumikoShirai,
Junichi Yoshida, Mike Barker
[2] Predicting Defect Densities in Source Code Files
with Decision Tree Learners Patrick Knab, Martin
Pinzger, Abraham Bernstein
ISSN: 2231-5381
http://www.ijettjournal.org
Page 11
International Journal of Engineering Trends and Technology (IJETT) – Volume 22 Number 1- April 2015
[3] Comparative Assessment of Software Quality
Classification Techniques: An Empirical Case Study
TAGHI M. KHOSHGOFTAAR, NAEEM SELIYA
[4] A.S. Foulkes, Applied Statistical Genetics with R.
Springer, 2009.
[5] A.L. Goel and K. Okumoto, “Time-Dependent
Error-Detection Rate Model for Software Reliability
and Other Performance Measures,” IEEE Trans.
Reliability, vol. 28, no. 3, pp. 206-211,Aug. 1979
[6] T. Ball, “A theory of predicate-complete test
coverage and generation,” inFMCO, 2004, pp. 1–22.
[7] V. R. Basili, L. C. Briand, and W. L. Melo, “A
validation of object oriented design metrics as quality
indicators,”IEEE Trans. Softw. Eng., vol. 22, no. 10,
pp. 751–761, 1996.
[8] R. P. L. BuseandW.Weimer, “Automatic
documentation inference for exceptions,” in ISSTA,
2008, pp. 273–282.
[9] ——, “A metric for software readability,” in
International Symposium on Software Testing and
Analysis, 2008, pp. 121–130.
[10] ——, “The road not taken: Estimating path
execution frequency statically,” in ICSE, 2009, pp.
144–154.
[11] H. Chen, D. Wagner, and D. Dean, “Setuid
demystified,” in USENIX Security Symposium, 2002,
pp. 171–190.
[12] S. R. Chidamber and C. F. Kemerer, “A metrics
suite for object oriented design,”IEEE Trans. Softw.
Eng., vol. 20, no. 6, pp. 476– 493, 1994.
ISSN: 2231-5381
http://www.ijettjournal.org
Page 12