M. Sadreddini

Followers

Following

Co-authors

Public Views

Interests

Uploads

Papers by M. Sadreddini

IQPI: An Incremental System for Answering Imprecise Queries Using Approximate Dependencies and Concept Similarities

1Abstract—Most of the proposed systems to process queries over web databases require the user to ... more 1Abstract—Most of the proposed systems to process queries over web databases require the user to provide some information regarding the relative importance of attributes and the similarities between nominal values. Recently, a new system called AIMQ has been proposed, which is based on measuring concept similarities. This system is end-user independent and can answer imprecise queries. The main drawback of this system is that it is not incremental. All computations must be repeated when a tuple is added to the database. As a solution to this problem, in this article, we propose an incremental and efficient system called IQPI, which can be considered as the incremental version of AIMQ. In IQPI, the set of approximate dependencies between attributes are mined, first (using our new efficient approach). Using this set of dependencies, the user's imprecise query is converted into some precise queries. Each of the precise queries is then fed into the system and the results are filter...

Download

Mining Fuzzy Association Rules Using Mutual

Quantitative Association Rule (QAR) mining has been recognized as an influential research problem... more Quantitative Association Rule (QAR) mining has been recognized as an influential research problem over the last decade due to the popularity of quantitative databases and the usefulness of association rules in real life. However, the combination of these quantitative attributes and their value intervals always rise to the generation of an explosively large number of itemsets, thereby severely degrading the mining efficiency. In this paper, we introduce a novel technique, called MFAMI, for mining quantitative association rules using fuzzy set theory. MFAMI employs linguistic terms to represent the revealed regularities and exceptions. This algorithm avoid the costly generation of a large number of candidate sets. Instead, using mutual information indicates the strong informative relationships among the attributes; potential frequent itemsets will be discovered. By utilizing those itemsets we devise an efficient algorithm that significantly generates rules. For effective mine rules, MFAMI employs adjusted difference analysis with this advantage that it does not require any user-supplied thresholds which are often hard to determine. Since the proposed algorithm greatly reduces the candidate subsequence generation efforts, the performance is improved significantly. Experiments show that the proposed algorithm is capable of discovering meaningful and useful fuzzy association rules in an effective manner, speeding up the mining process and obtaining most of the high confidence QARs.

Download

PTclose: A novel algorithm for generation of closed frequent itemsets from dense and sparse datasets

In recent years, various algorithms for mining closed frequent itemsets (CFI) have been proposed.... more In recent years, various algorithms for mining closed frequent itemsets (CFI) have been proposed. Different structures such as prefix sharing are used within these algorithms. However the name challenging problem in many of them is the high requirement of memory, especially in case of sparse datasets. Thus, most of the proposed methods are proper for dense datasets. In this paper we present a new approach to mining closed frequent itemsets using two structures, namely Patricia tree and PTArray. By using these two structures, both the response time and also memory consumption are reduced, significantly. The proposed method, called PTclose is suitable for both dense and sparse datasets. The algorithm is assessed throughout a set of experiments. The results narrate for the relative efficiency of the algorithm compared to other existing methods.

Download

Using Candidate Hashing and Transaction Trimming in Distributed Frequent Itemset Mining

Abstract: Frequent itemset mining is one of the most popular data mining tasks which has wide app... more Abstract: Frequent itemset mining is one of the most popular data mining tasks which has wide application areas and is redefined in distributed environment. Computation and communication are two important factors in distributed frequent itemset mining. In this paper two techniques have been exploited to reduce communication and improve the running time in a distributed environment. These techniques were proposed previously for centralized setting and here they are adopted in the FDM algorithm as one of the well-known distributed association rules mining algorithm. The proposed algorithm uses Trie data structure for better performance. Experimental evaluations on different sort of distributed data show the effect of using these adopted techniques.

Download

Imprecise Queries Using Approximate Dependencies and Concept Similarities

Abstract—Most of the proposed systems to process queries over web databases require the user to p... more Abstract—Most of the proposed systems to process queries over web databases require the user to provide some information regarding the relative importance of attributes and the similarities between nominal values. Recently, a new system called AIMQ has been proposed, which is based on measuring concept similarities. This system is end-user independent and can answer imprecise queries. The main drawback of this system is that it is not incremental. All computations must be repeated when a tuple is added to the database. As a solution to this problem, in this article, we propose an incremental and efficient system called IQPI, which can be considered as the incremental version of AIMQ. In IQPI, the set of approximate dependencies between attributes are mined, first (using our new efficient approach). Using this set of dependencies, the user's imprecise query is converted into some precise queries. Each of the precise queries is then fed into the system and the results are filtere...

Download

Combining Different Seed Dictionaries to Extract Lexicon from Comparable Corpus

Indian Journal of Science and Technology

In recent years, many studies on extracting new bilingual lexicons from non-parallel (comparable)... more In recent years, many studies on extracting new bilingual lexicons from non-parallel (comparable) corpora have been proposed. Nearly all apply an existing small dictionary or other resource to make an initial list named seed dictionary. In this paper we discuss on using different types of dictionaries and their combinations as the initial starting list to produce a bilingual Persian-Italian lexicon from a comparable corpus. Our experiments applied state of the art techniques on four different seed dictionaries; an existing dictionary and three dictionaries created with pivot-based schema considering three different languages as pivot. We have used English, Arabic and French as pivot languages to extract these three pivot based dictionaries. An interesting challenge in our approach is proposing a method to combine different dictionaries together producing a better and more accurate lexicon. In order to combine seed dictionaries we proposed two novel combination models and examine the effect of them on comparable corpora which are collected from News Agencies. The experimental results exploited by our implementation show the efficiency of our proposed combinations.

Download

Frequent Patterns Mining over Data Stream Using an Efficient Tree Structure

Mining frequent patterns over data streams is an interesting problem due to its wide application ... more Mining frequent patterns over data streams is an interesting problem due to its wide application area. In this study, a novel method for sliding window frequent patterns mining over data streams is proposed. This method utilizes a compressed and memory efficient tree data structure to store and to maintain sliding window transactions. The method dynamically reconstructs and compresses tree data structure to control the amount of memory usage. Moreover, the mining task is efficiently performed using the data structure when a user issues a mining request. The mining process reuses the tree structure to extract frequent patterns and does not need additional memory requirement. Experimental evaluations on real datasets show that our proposed method outperforms recently proposed sliding window based algorithms.

Download

A framework for integrating distributed healthcare information systems

International Journal of Information Science and Management, 2012

In the last two decades, the trend in healthcare information technology has been towards digital ... more In the last two decades, the trend in healthcare information technology has been towards digital and multimedia concepts. The next generation of healthcare information systems will consist of a vast network of heterogeneous, autonomous and distributed imaging scanners, Hospital Information Systems (HIS), medical information from clinical departments such as Radiology Information Systems (RIS), and large quantities of multimedia medical data held on a range of Picture Archiving and Communication Systems (PACS). Providing an organizational framework that can integrate this varied collection of resources is a key challenge system researchers and developers face. In order to increase the availability of global or previously non-accessible information and to address the demanding new information processing requirements for diverse image-assisted medical applications, there is a need to integrate HIS/RIS data and medical image data of Picture Archiving and Communication Systems (PACS). Th...

Applying Weighted KNN to Word Sense Disambiguation

One of the major challenges in the process of machine translation is word sense disambiguation (W... more One of the major challenges in the process of machine translation is word sense disambiguation (WSD), which is defined as choosing the correct meaning of a multimeaning word. Supervised learning methods are usually used to solve this problem. The disambiguation task is carried out using the statistics of the translated documents (as training data) or dual corpora of source and target languages. In this paper we present a supervised learning method for WSD, which is based on K-Nearest Neighbor algorithm. As the first step, we extract two sets of features; the set of words that have occurred frequently in the text and the set of words surrounding the ambiguous word. In order to improve the classification accuracy, we propose a feature weighting strategy. We will present the results of evaluating the proposed schemes and illustrate the effect of weighting strategies proposed. The results are encouraging comparing to state of the art.

Download

Machine Translation Based on Data Mining and Deductive Schemes

Machine translation (MT) is one of the most attractive fields in natural language processing. In ... more Machine translation (MT) is one of the most attractive fields in natural language processing. In this paper, we propose some new ideas for designing an MT system. For this purpose, we first introduce a grammatical rule induction method. After representing the extracted knowledge by a set of finite automata, a recursive model is proposed, which uses a combination of rule and example based techniques. In the translation phase, through a hierarchical chunking process, the input sentence is divided into a set of phrases. Each phrase is first searched through the corpus of examples. If the phrase is found, it will not be chunked anymore. Otherwise, the phrase is divided into smaller sub-phrases. The experimental results show a promising accuracy and efficiency of the proposed system.

Download

AD-Miner: A new incremental method for discovery of minimal approximate dependencies using logical operations

Intelligent Data Analysis

ABSTRACT

EclatDS: An efficient sliding window based frequent pattern mining method for data streams

Intelligent Data Analysis

... streams Mahmood Deypir∗ and Mohammad Hadi Sadreddini Department of Computer Science and Engin... more

Implementation of Learning Management System Model in Vocational School

Proceedings of the 2015 International Conference on Innovation in Engineering and Vocational Education, 2016

This study aims to develop and to implement the Learning Management System (LMS) model in improvi... more This study aims to develop and to implement the Learning Management System (LMS) model in improving learning outcomes in Electronic Measurement (EM) subjects for students of vocational school in Bandung. LMS is an online learning system or e-learning, which is a web-based learning management application that can facilitate students/learners to learn in a virtual classroom. This research employed research and development approach, which led to a cycle based on studies and research findings were then developed to produce a model of the product. To find an increase in students' learning outcomes in the cognitive aspects, the quantitative approach was done with a technique that used a quasi-experimental pretest-posttest group design. The experimental group was given a learning treatment with LMS implementation, while in the control group was treated with conventional learning models approach in the classroom. The data obtained in this study were from the results of the pretest and posttest and then analyzed with an average ratio test (t-test) and test score to gain the difference. The results of this study indicated that vocational students' achievement in EM subject by using LMS applications models were higher than the students' learning outcomes by using conventional learning models. Based on this analysis, it can be concluded that LMS model can improve students' learning outcomes in EM subject at vocational school in Bandung.

Download

Fuzzy Accuracy-based Classifier Systems for EEG Classification of Schizophrenic Patients

In this paper, EEG signals of twenty schizophrenic patients and twenty age-matched healthy subjec... more In this paper, EEG signals of twenty schizophrenic patients and twenty age-matched healthy subjects are analyzed with objective of classification two groups. In this case, Fuzzy Accuracy-based Classifier System (F-XCS) is used to automatically generate fuzzy if-then rules for discrimination healthy and schizophrenic subjects. Several features including AR model coefficients, band power and fractal dimension are extracted from EEG signals. First, the F-XCS is applied where a randomly generated initial population of fuzzy if-then rules is evolved by typical genetic operations such as selection, crossover and mutation. Second, a heuristic procedure for improving the performance of F-XCS is applied and result of adding this heuristic procedure is analyzed. The motivation behind this approach is that F-XCS will be capable of generating compact, high performance rule sets which are general and accurate.

Download

Mining Frequent Itemsets in Large Data Warehouses: A Novel Approach Proposed for Sparse Data Sets

Lecture Notes in Computer Science, 2007

Proposing efficient techniques for discovery of useful information and valuable knowledge from ve... more Proposing efficient techniques for discovery of useful information and valuable knowledge from very large databases and data warehouses has attracted the attention of many researchers in the field of data mining. The wellknown Association Rule Mining (ARM) algorithm, Apriori, searches for frequent itemsets (i.e., set of items with an acceptable support) by scanning the whole database repeatedly to count the frequency of each candidate itemset. Most of the methods proposed to improve the efficiency of the Apriori algorithm attempt to count the frequency of each itemset without re-scanning the database. However, these methods rarely propose any solution to reduce the complexity of the inevitable enumerations that are inherited within the problem. In this paper, we propose a new algorithm for mining frequent itemsets and also association rules. The algorithm computes the frequency of itemsets in an efficient manner. Only a single scan of the database is required in this algorithm. The data is encoded into a compressed form and stored in main memory within a suitable data structure. The proposed algorithm works in an iterative manner, and in each iteration, the time required to measure the frequency of an itemset is reduced further (i.e., checking the frequency of ndimensional candidate itemsets is much faster than those of n-1 dimensions). The efficiency of our algorithm is evaluated using artificial and real-life datasets. Experimental results indicate that our algorithm is more efficient than existing algorithms.

Download

A Robust Instance Weighting Technique for Nearest Neighbor Classification in Noisy Environments

Indian Journal of Science and Technology, 2015

The performance of Nearest Neighbor (NN) classifier is highly dependent on the distance (or simil... more The performance of Nearest Neighbor (NN) classifier is highly dependent on the distance (or similarity) function used to find the NN of an input test pattern. Many of the proposed algorithms try to optimize the accuracy of the NN rule using a weighted distance function. In this scheme, a weight parameter is learned for each of the training instances. The weights of training instances are used in the generalization phase to find the NN of an input test pattern. The Weighted Distance Nearest Neighbor (WDNN) algorithm attempts to maximize the leave-one-out classification rate of the training set by adjusting the weight parameters. The procedure simply leads to weights that overfit the train data, which degrades the performance of the method especially in noisy environments. In this paper, we propose an enhanced version of WDNN, called Overfit Avoidance for WDNN (OAWDNN), that significantly outperforms the algorithm in generalization phase. The proposed method uses an early stopping approach to decrease instance weights specified by WDNN, which implicitly makes the class boundary smooth and consequently more generalized. In order to evaluate robustness of the algorithm, class label noise is added to a variety of UCI datasets. The experimental results show the supremacy of the proposed method in generalization accuracy.

<title>Architectural considerations for integrating HIS/RIS/PACS in a distributed and heterogeneous environment</title>

Medical Imaging 1993: PACS Design and Evaluation, 1993

Advances in image processing techniques, databases, workstations, network technologies and commun... more Advances in image processing techniques, databases, workstations, network technologies and communication systems have created great interest in the integration of heterogeneous collections of multimedia data. The success of a medical multimedia information system depends on integration of distributed heterogeneous PACS, RIS and HIS, to provide rapid access to patients' text and image information throughout a hospital. The work presented here adapts a distributed database (DDB) approach to integrating HIS, RIS and PACS in a distributed heterogeneous environment. It is based on previous work on integrating structured (relational) database systems using the `Multidatabase' approach. Integrating distributed heterogeneous HIS/RIS/PACS causes special problems. Two alternative architectural proposals for providing system support to integrate heterogeneous HIS/RIS and PACS in a distributed environment are presented and discussed. One of these architectures provides system support by using a distributed index approach, where the images indexes are held at the local sites. The second proposed architecture uses a global image index. In this approach the global image index provides the list of available images within the network. Advantages and disadvantages of each approach are also discussed.

Mining Frequent Itemsets in Distributed Environment

Lecture Notes in Electrical Engineering, 2009

Finding association rules is one of the most investigated fields of data mining. Computation and ... more Finding association rules is one of the most investigated fields of data mining. Computation and communication are two important factors in distributed association rule mining. In this problem Association rules are generated by first mining of frequent itemsets in distributed data. In this paper we proposed a new distributed trie-based algorithm (DTFIM) to find frequent itemsets. This algorithm is proposed

Evaluating different clustering techniques for electricity customer classification

IEEE PES T&D 2010, 2010

... 105-138. VII. BIOGRAPHIES Seyed Mohammad Bidoki was born in Yazd, May 26, 1985. He received C... more

A new word sense disambiguation system based on deduction

Word sense ambiguity resolution is one of the major issues in the process of machine translation.... more Word sense ambiguity resolution is one of the major issues in the process of machine translation. Statistical and example-based methods are usually applied for this purpose. In statistical methods, ambiguity resolution is mostly carried out by making use of some statistics extracted from previously translated documents or dual corpora of source and target languages. In this paper, we look at the problem from a different viewpoint. The proposed system consists of two main parts. The first part includes a data mining algorithm which runs offline and extracts some useful knowledge about the cooccurrences of the words. The second part of the system is an expert system whose knowledge base includes the set of association rules generated by the first part. For the inference engine of the expert system, we propose an efficient algorithm based on forward chaining in order to deduce the correct senses of the words. The performance of the system in terms of applicability and precision will be analyzed and discussed through a set of experiments.

Download

M. Sadreddini

Uploads

Papers by M. Sadreddini

Log In