Welcome to Journal of Theoretical and Applied Information Technology

Language is a unique phenomenon that distinguishes man from other animals. It is our primary method of communication with each other, yet very little is understood about how language is acquired when we are infants. A greater understanding in this area would have the potential to improve man machine communication. The problem that is attempted to be solved in this paper is that of programming a computer to play the Shannon Game. To play the Shannon game, one must predict which words are most likely to follow a given segment of English Text. Word Prediction would be most useful for writers with physical disabilities and severe spelling problems. The aim of this paper is to improve on existing results by writing a program that is capable of automatically inferring a grammar from a Natural Language Corpus, and applying this to the Shannon Game. To play the Shannon Game, a stochastic Grammar for an approximation to the target language must be inferred from a text sample, and as the quality of this grammar improves so too does the quality of the predictor that uses the inferred grammar. The proposed algorithm in the paper uses Support Vector Machine to perform the part of speech tagging which produces 97.6% correct predictions.

Full Text

A MODEL FOR OVERLAPPING TRIGRAM TECHNIQUE FOR TELUGU SCRIPT

B.Vishnu Vardhan, L.Pratap Reddy, A.VinayBabu

N-grams are consecutive overlapping N-character sequences formed from an input stream. N-grams are used as alternatives to word-based retrieval in a number of systems. In this paper we propose a model applicable to categorization of Telugu document. Telugu is an official language derived from ancient Brahmi script and also the official language of the state of Andhra Pradesh. Brahmi based script is noted for complex conjunct formations. The canonical structure is described as ((C) C) CV. The structure evolves any character from a set of basic syllables known as vowels and consonants where consonant, vowel (CV) core is the basic unit optionally preceded by one or two consonants. A huge set of characters that resemble the phonetic nature with an equivalent character shape are derived from the canonical structure. Words formed from this set evolved into a large corpus.  Stringent grammar rules in word formation are part of this corpus. Certain word combinations result in the formation of single word is to be addressed where the last character of the first word and first character of the successive word are combined. Keeping in view of these complexities we propose a trigram based system that provides a reasonable alternative to a word based system in achieving document categorization for the language Telugu.

APPLICATION OF KNOWLEDGE MANAGEMENT IN MANAGEMENT EDUCATION:
A CONCEPTUAL FRAMEWORK

Jayanthi Ranjan, Saani Khalil

The paper presents a conceptual framework in the context of Knowledge Management (KM) in Business Schools (B-schools) in India. We believe that if the framework is adopted in business schools, it will yield more benefits to increase the quality of knowledge sharing. There has been indeed a paradigm shift in management education in India. The new breed of management professionals need to be efficient to tackle problems from cross functional, cultural and ethical perspectives and equipped with skills to bench mark for global leadership positions. There has been a crying need to usher in a quality movement and to benchmark the same with world standards. We have made an attempt to support our framework by analyzing one of the Knowledge Management tools that was implemented in India’s Test Institute of Management (TIM), (a pseudonym is given to mask the institution’s name). This paper studies the knowledge management tool and features that are implemented in TIM and some problems that hindered knowledge management practices at TIM.

Full Text

Sibaram Khara, Asish K. Mukhopadhyay

This paper proposes a new technique for Mobile IP (MIP) registration by WLAN host (WH) of the Internet through GPRS network. The home agent (HA) of WH resides in the Internet. The gateway GPRS Support Node (GGSN) provides foreign agent (FA) functionality for WH in GPRS. After successful attachment with GPRS network in roaming scenario, WH needs two mandatory passes for MIP registration with HA. In first pass, it establishes PDP context in GPRS network and in second pass, it sends MIP registration request to FA at GGSN. This two-pass registration causes substantial delay for handoff from WLAN to GPRS. To minimize this delay, we propose a one-pass MIP registration technique through GPRS network by modifying the activate-PDP-context request (APCR) message. In this scheme, the APCR message carries MIP-registration request in its information field to GGSN. Therefore, FA at GGSN can process MIP registration packet before completion of PDP context in GPRS network. Thus, both PDP context creation and MIP registration are accomplished in one pass of signaling. This technique reduces handoff delay from WLAN to GPRS minimizing the control signals. We observed from simulation results that the proposed one-pass technique reduces handoff delay by 18% compared to that of two-pass method proposed in [8].

In this paper, we propose two approach intelligent techniques of improvement of Direct Torque Control (DTC) of Induction motor such as fuzzy logic (FL) and artificial neural network (ANN), applied in switching select voltage vector .The comparison with conventional direct torque control (DTC), show that the use of the DTC_FL and DTC_ANN, reduced the torque, stator flux, and current ripples. The validity of the proposed methods is confirmed by the simulative results.

Full Text

ANN BASED CONTROL PATTERNS ESTIMATOR FOR UPFC USED IN POWER FLOW PROBLEM

K.Krishnaveni, G. Tulasi Ram Das

The continuous growth in the demand for electric power necessitates the flexibility of operation in power system. Of different power electronics-based Flexible AC Transmission System (FACTS) devices, which enhance the power transmission capabilities, Unified Power Flow Controller (UPFC) provides an emerging and promising solution for the power flow problems in the system, as it simultaneously and/or selectively controls the transmission parameters. In this context, the paper proposes the power flow control in a simple system by injecting the series compensating voltage, which is an important function of UPFC. For this purpose, ANN controller based UPFC is used. Control patterns are generated for obtaining the adjustable series voltage from the second converter that, in turn, controls the power flow in the system. With the proposed model, by varying control coefficient the series injected voltage can be adjusted. MATLAB Simulation is used to test the proposed model. The control horizon is identified and presented for various values of existing active and reactive powers.

Full Text

PERFORMANCE OF ANN BASED INDIRECT VECTOR CONTROL INDUCTION MOTOR
DRIVE

A. K. Sharma, R. A. Gupta, Laxmi Srivastava

Indirect field orientation (IFO) induction machine drives are increasingly employed in industrial drive systems, but the drive performance is often degrades. Motor works on best performance at certain voltage and frequency for certain loads. In this paper artificial neural network is used to predict the operating voltage and frequency when the load torque and speed going changed so motor efficiency is increased. Simulation and experimental results are shown to validate the scheme.

Full Text

Distributed Dynamic Channel Allocation Algorithm for Cellular Mobile Network

Megha Gupta, A.K. Sachan

A channel allocation algorithm includes a channel acquisition and a channel selection scheme. Most of the previous work concentrates on the channel selection algorithm since early channel acquisition algorithms are centralized and rely on a MSS to accomplish channel acquisition. The centralized schemes are neither scalable nor reliable. Recently, distributed dynamic channel allocation algorithms have received considerable attention due to their high reliability and scalability. The most of the distributed algorithm is based on non-resource planning model in which a borrower needs to consult with every interference neighbors in order to borrow a channel. The proposed distributed dynamic channel allocation algorithm is based on resource-planning model, a borrower need not to receive replies from every interfering neighbors, it can borrow a channel from that neighbor whose all group members replies with common free channels within the predefined time period. The proposed algorithm makes efficient reuse of channels and evaluates the performance in terms of message complexity, blocking rate.

Full Text

NOVEL CONSTRUCTION OF SHORT LENGTH LDPC CODES FOR SIMPLE DECODING

This paper introduces different construction techniques of parity-check matrix H for irregular Low-Density Parity-Check (LDPC) codes. The first one is the proposed Accurate Random Construction Technique (ARCT) which is an improvement of the Random Construction Technique (RCT) to satisfy an accurate profile. The second technique, Speed Up Technique (SUT), improves the performance of irregular LDPC codes by growing H from proposed initial construction but not from empty matrix as usual. The third and fourth techniques are further improvements of the SUT that insure simpler decoding. In Double Speed Up Technique (DSUT), the decoder size of SUT matrix is fixed and the size of H is doubled. In Partitioned Speed Up Technique (PSUT), the H size is fixed and the decoder size decreases by using small size of SUT matrices to grow H. Simulations show that the performance of LDPC codes formed using SUT outperforms ARCT at block length N = 1000 with 0.342dB at BER = 10-5 and LDPC codes created by DSUT outperforms SUT with 0.194dB at BER = 10-5. Simulations illustrate that the partitioning of H to small SUT submatrices not only simplifies the decoding process, it also simplifies the implementation and improves the performance. The improvement, in case of half, is 0.139dB at BER=10-5 however as partitioning increases the performance degrades. It is about 0.322dB at BER=10-5 in case of one-fourth.

Full Text

INTEGRATION OF SELF ORGANIZING FEATURE MAPS AND HONEY BEE MATING OPTIMIZATION ALGORITHM FOR MARKET SEGMENTATION

Babak Amiri, Mohammad Fathian

This study is dedicated to proposing a two-stage method, which first uses Self-Organizing Feature Maps (SOM) neural network to determine the number of clusters and cluster centroids, then uses honey bee mating optimization algorithm based on K-means algorithm to find the final solution. The results of simulated data via a Monte Carlo study show that the proposed method outperforms two other methods, SOM followed by K-means (Kuo, Ho & Hu, 2002a) and SOM followed by GAK (Kuo, An, Wang & Chung, 2006), based on both within-cluster variations (SSW) and the number of misclassification. In order to further demonstrate the proposed approach’s capability, a real-world problem of an internet bookstore market segmentation based on customer loyalty is employed. The RFM model is used for comparison of customers' loyalty. Then the proposed method is used to cluster the customers. The results also indicate that the proposed method is better than the other two methods.

Full Text