Occurrence Matrix Research Articles

A novel statistical feature extraction method for textual images: Optical font recognition

The binary image is essential to image formats where the textual image is the best example of the binary image representation. Feature extraction is a fundamental process in pattern recognition. In this regard, pattern recognition studies involve document analysis techniques. Optical font recognition is among the pattern recognition techniques that are becoming popular today. In this paper, we propose an enhanced global feature extraction method based on the on statistical analysis of the behavior of edge pixels in binary images. A novel method in feature extraction for binary images has been proposed whereby the behavior of the edge pixels between a white background and a black pattern in a binary image captures information about the properties of the pattern. The proposed method is tested on an Arabic calligraphic script image for an optical font recognition application. To evaluate the performance of our proposed method, we compared it with a gray-level co occurrence matrix (GLCM). We classified the features using a multilayer artificial immune system, a Bayesian network, decision table rules, a decision tree, and a multilayer network to identify which approach is most suitable for our proposed method. The results of the experiments show that the proposed method with a decision tree classifier can boost the overall performance of optical font recognition.

자궁경부암 진단을 위한 3차원 세포핵 질감 특성값 유의성 평가에 관한 연구

본 연구의 목적은 세포핵의 3차원 염색질 질감 특성값이 암의 진행정도를 인식하는데 있어 유용한 특성값인지 평가하는데 있다. 특히, 제안한 방법이 악성이라고 진단된 세포진 도말 표본에서 정상으로 보이는 세포의 염색질 패턴에서의 미세한 차이를 인식할 수 있는지 살펴보고자 한다. 분류등급 정상(Normal), 저등급 편평 상피내 병변(LSIL, Low grade Squamous Intraepithelial Lesion), 고등급 편평 상피내 병변(HSIL, High grade Squamous Intraepithelial Lesion)에서 각각 100개씩의 세포 볼륨데이터로부터 3차원 GLCM(Gray Level Co occurrence Matrix)에 기반한 질감 특성값과 3차원 Wavelet 변환에 기반한 질감 특성값을 추출하고 분류기를 생성한 후 각 분류기에 대한 분류정확도를 비교하였으며, 2차원 세포진 영상에서의 세포핵 질감 특성값과 비교하기 위해 동일한 실험 볼륨데이터의 투영된 2차원 영상을 이용하여 같은 방법으로 2차원 세포핵 질감 특성값을 추출하고 분류기를 생성한 후 분류정확도를 비교하였다. 2차원 세포핵 질감 특성값과의 비교연구에서 3차원 세포핵 질감 특성값이 등급별 분류에 있어 보다 효율적인 것을 확인 할 수 있었으며 이는 3차원 염색질 질감 특성값이 자궁경부 세포의 정량화에 대한 정확성과 재현성을 개선할 수 있음을 의미한다. The aim of this study is to evaluate whether 3D nuclear chromatin texture features are significant in recognizing the progression of cervical cancer. In particular, we assessed that our method could detect subtle differences in the chromatin pattern of seemingly normal cells on specimens with malignancy. We extracted nuclear texture features based on 3D GLCM(Gray Level Co occurrence Matrix) and 3D Wavelet transform from 100 cell volume data for each group (Normal, LSIL and HSIL). To evaluate the feasibility of 3D chromatin texture analysis, we compared the correct classification rate for each of the classifiers using them. In addition to this, we compared the correct classification rates for the classifiers using the proposed 3D nuclear texture features and the 2D nuclear texture features which were extracted in the same way. The results showed that the classifier using the 3D nuclear texture features provided better results. This means our method could improve the accuracy and reproducibility of quantification of cervical cell.

텍스트마이닝 기반의 인적재난사고사례 신뢰도 측정연구

Deducting the risk level of infrastructure and buildings based on past human disaster risk cases and implementing prevention measures are important activities for disaster prevention. The object of this study is to measure the confidence to proceed quantitative analysis of various disaster risk cases through text mining methodology. Indeed, by examining confidence calculation process and method, this study suggests also a basic quantitative framework. The framework to measure the confidence is composed into four stages. First step describes correlation by categorizing basic elements based on human disaster ontology. Secondly, terms and cases of Term-Document Matrix will be created and the frequency of certain cases and terms will be quantified, the correlation value will be added to the missing values. In the third stage, association rules will be created according to the basic elements of human disaster risk cases. Lastly, the confidence value of disaster risk cases will be measured through association rules. This kind of confidence value will become a key element when deciding a risk level of a new disaster risk, followed up by preventive measures. Through collection of human disaster risk cases related to road infrastructure, this study will demonstrate a case where the four steps of the quantitative framework and process had been actually used for verification.

Integrating Document Clustering and Multidocument Summarization

Document understanding techniques such as document clustering and multidocument summarization have been receiving much attention recently. Current document clustering methods usually represent the given collection of documents as a document-term matrix and then conduct the clustering process. Although many of these clustering methods can group the documents effectively, it is still hard for people to capture the meaning of the documents since there is no satisfactory interpretation for each document cluster. A straightforward solution is to first cluster the documents and then summarize each document cluster using summarization methods. However, most of the current summarization methods are solely based on the sentence-term matrix and ignore the context dependence of the sentences. As a result, the generated summaries lack guidance from the document clusters. In this article, we propose a new language model to simultaneously cluster and summarize documents by making use of both the document-term and sentence-term matrices. By utilizing the mutual influence of document clustering and summarization, our method makes; (1) a better document clustering method with more meaningful interpretation; and (2) an effective document summarization method with guidance from document clustering. Experimental results on various document datasets show the effectiveness of our proposed method and the high interpretability of the generated summaries.

Query based Text Document Clustering using its Hypernymy relation

of text can be organized in an unsupervised manner. In this paper, Text document clustering is done based on query and its semantic relation. The method utilizes hypernymy to identify its relation. It was detected by using the Word Net. It act as background knowledge of the Query and provides its synonymic terms. This paper proposed the new term-document matrix called Query based document vector model, which is constructed using query with two terms and its hypernymy. The results show that our new measure Cluster Accuracy is significantly better to evaluate the quality of cluster and better results are obtained. General Terms Text mining and Information Retrieval, Partitioning M ethod KeywordsNoun, Word net, Query based document vector model, Hypernymy, Accuracy.

On Scale Invariance Texture Image Retrieval using Fuzzy Logic and Wavelet Co-occurrence based Features

this paper, analysis of the feature selection for scale invariance texture image retrieval using fuzzy logic classifier and wavelet and co-occurrence matrix based feature is carried out. Two types of texture features are extracted one using Discrete Wavelet Transform (DWT) and other using Co- occurrence matrix. Energy and Standard Deviation are obtained from each sub-band of DWT coefficients up to fifth level of decomposition and eight features are extracted from co-occurrence matrix of whole image and each sub-band of first level DWT decomposition. The different size samples of texture image are undertaken. The suitability of features extracted is analyzed using a fuzzy logic classifier. The performance is measured in terms of Success Rate. Best and Worst case analysis is done for each of the feature set and texture image size. Also the minimum number of features required for maximum average success rate is obtained. This study shows that for samples taken from 256x256 texture size, excellent success rate is achieved for Wavelet Statistical Features (WSF) as well as Wavelet Co-occurrence Features (WCF). Also WSF perform better for 128x128 and 256x256 texture image. For both the types of features performance degrades in case of 512x512 texture image. Worst case analysis shows that energy feature WSF and 8-features group WCF performs excellently.

CTDCIRS: Content based Image Retrieval System based on Dominant Color and Texture Features

There is a great need of developing efficient content based image retrieval systems because of the availability of large image databases. A new image retrieval system CTDCIRS (colortexture and dominant color based image retri eval system) to retrieve the images using three features called dynamic dominant color (DDC), Motif cooccurrence m atrix (MCM) and difference between pixels of scan pattern (DBPSP) is proposed. Initially the image is divided into eight coarse partitions using the fast color quantization algorithm and the eight dominant colors are obtained from eight partitions. Next the texture of the image is represented by the MCM and DBPSP. MCM is derived using a motif transformed image. MCM is similar to color cooccur rence matrix (CCM). MCM is the conventional pattern co occurrence matrix that calculates the probability of the occurrence of same pixel color between each pixel and its adjacent ones in each image, and this probability is considered as the attribute of the image.MCM captures third order image statistics in the local neighborhood which describes the direction of textures but not the complexity of the textures. That is why the DBPSP is also considered as one of the texture features. The three features Dominant color, MCM and DBPSP are integrated to facilitate the image retrieval system. Experimental results show that the proposed image retrieval is more efficient in retrieving the user interested images.

LSA based multi-instance learning algorithm for image retrieval

Focusing on the problem of natural image retrieval, based on latent semantic analysis (LSA) and support vector machine (SVM), a novel multi-instance learning (MIL) algorithm is proposed, where a bag corresponds to an image and an instance corresponds to the low-level visual features of a segmented region. Firstly, in order to transform every bag into a single sample, a collection of “ visual-word” is generated by k-means clustering method to construct a projection space, then a nonlinear mapping is defined using these “ visual-word” to embed each bag as a point in the projection space, thereby obtaining every bag's projection feature. Secondly, the matrix consisted of all the projection features of training bags is regarded as a term-document matrix, and LSA method is used to obtain the latent semantic feature of each bag. As a result, the MIL problem is converted into a standard single instance learning (SIL) problem that can be solved directly by SVM method. Experimental results on the COREL data sets show that the proposed method, named LSASVM-MIL, is robust, and its performance is superior to other key existing MIL algorithms.

Benthic Classifications Using Bathymetric LIDAR Waveforms and Integration of Local Spatial Statistics and Textural Features

The scope of this research is to assess benthoscape discrimination by airborne light detection and ranging (LIDAR) bathymetry (ALB) on the basis of statistical parameters derived from the LIDAR waveforms, textural information, and local spatial statistics. Analysis of the underwater camera stations allowed clustering of the stations into groups on the basis of their habitat composition (β-diversity). Twelve descriptive statistics describing the shape of the bottom part of the waveform, also called 12 benthic parameters, were used for discriminating four benthic habitats. A K-means classification and a supervised method based on the Support Vector Machine (SVM) were applied to this dataset, and overall accuracies of 67.7% and 89.9% were obtained, respectively. Geostatistical analyses, using 11 textural measures, defined by the gray-level occurrence matrix (GLOM) and the gray-level co-occurrence matrix (GLCM), and three local spatial statistics were then applied to the 12 benthic parameters to enhance the SVM classification performance. The assessment of the contribution of geostatistics into benthic class segmentation was achieved by computation of separability distance. Mean (from the GLOM), mean (from the GLCM) and the local Getis-Ord statistic yielded the best rates of discrimination. These added metrics, integrated with bands related to the 12 benthic parameters, showed that the rate of correct (supervised) classification was thereby improved and increased by 5.3%. Finally, the first four principal components (PCs) (i.e., 90.41% of the 12 parameter variances, boosted by the three best geostatistics) brought out an overall accuracy of 93.3%, showing evidence for optimizing the classification processing.

Topic Mining based on Word Posterior Probability in Spoken Document

For speech recognition system, there are three kinds of result representations as one-best, N -best and Lattice. Since lattice has multi-path which can reduce the effect of recognition error rate, it is widely applied nowadays. In fact, there are amount of redundancies in lattice, which leads to the increasing of complexity of latter algorithm based on it. Additionally, for the decoding algorithm, it is acted as maximum a posterior probability (MAP) which can only guarantee the posterior probability of the whole sentence is of maximum. For MAP does not mean the highest syllable recognition rate, here, confusion network is introduced in topic mining system. In the clustering during confusion network, the minimum word error rule is adopted, which is proper to topic mining system since the least meaningful unit is word in Chinese and word information is most important in topic mining. In this paper, a simplified confusion network generation algorithm is proposed to handle some problems caused by insertion error during recognition. Then based on the confusion network, a word list extraction approach is proposed, in which, the dictionary is adopted to judge whether the consecutive arc in confusion sets is a word. At this stage, the error word information produced by error recognition rate can be corrected to some extent. After the competition part in word list extraction on confusion network, a final word list with posterior probability can be obtained. Furthermore, this kind of posterior probability can be combined in topic mining system. SVD and NMF are adopted here to decompose the term-document matrix on the word list of confusion network. From the experiments, it can be drawn that the proposed approach based on confusion network can achieve better performance than that of one-best and N -best. Additionally, the modified weight which combined posterior probability into term-document matrix can further improve the system performance.

Hybrid index-based image search from the web

Existing search techniques for retrieving images from the web store text-based and content-based features separately. They use structures like inverted-index, forward-index, document-term matrix, Tries, Prefix B-Tree, String B-Tree, etc. for text-based features and R-tree, SR-tree, K-B-D Tree, etc., for content-based features. We propose to use a hybrid indexing scheme which is more intuitive for hybrid image feature vectors and can be used to both store and query non-ordered discrete and continuous features simultaneously. Also, since most of the existing hybrid image search engines do not store two types of features together, they usually perform retrieval in two distinct steps, first finding results with only text-based information and later filtering results based on content-based information. In contrast, our approach of hybrid indexing supports retrieval in a single step. We introduce a k-nearest neighbour search algorithm for the hybrid indexing scheme used.

Evaluation of Co-occurring Terms in Clinical Documents Using Latent Semantic Indexing

ObjectivesMeasurement of similarities between documents is typically influenced by the sparseness of the term-document matrix employed. Latent semantic indexing (LSI) may improve the results of this type of analysis.MethodsIn this study, LSI was utilized in an attempt to reduce the term vector space of clinical documents and newspaper editorials.ResultsAfter applying LSI, document similarities were revealed more clearly in clinical documents than editorials. Clinical documents which can be characterized with co-occurring medical terms, various expressions for the same concepts, abbreviations, and typographical errors showed increased improvement with regards to a correlation between co-occurring terms and document similarities.ConclusionsOur results showed that LSI can be used effectively to measure similarities in clinical documents. In addition, correlation between the co-occurrence of terms and similarities realized in this study is an important positive feature associated with LSI.

Nonnegative Matrix Factorization with Orthogonality Constraints

Nonnegative matrix factorization (NMF) is a popular method for multivariate analysis of nonnegative data, which is to decompose a data matrix into a product of two factor matrices with all entries restricted to be nonnegative. NMF was shown to be useful in a task of clustering (especially document clustering), but in some cases NMF produces the results inappropriate to the clustering problems. In this paper, we present an algorithm for orthogonal nonnegative matrix factorization, where an orthogonality constraint is imposed on the nonnegative decomposition of a term-document matrix. The result of orthogonal NMF can be clearly interpreted for the clustering problems, and also the performance of clustering is usually better than that of the NMF. We develop multiplicative updates directly from true gradient on Stiefel manifold, whereas existing algorithms consider additive orthogonality constraints. Experiments on several different document data sets show our orthogonal NMF algorithms perform better in a task of clustering, compared to the standard NMF and an existing orthogonal NMF.

Relationships among cloud occurrence frequency, overlap, and effective thickness derived from CALIPSO and CloudSat merged cloud vertical profiles

A cloud frequency of occurrence matrix is generated using merged cloud vertical profiles derived from the satellite‐borne Cloud‐Aerosol Lidar with Orthogonal Polarization (CALIOP) and cloud profiling radar. The matrix contains vertical profiles of cloud occurrence frequency as a function of the uppermost cloud top. It is shown that the cloud fraction and uppermost cloud top vertical profiles can be related by a cloud overlap matrix when the correlation length of cloud occurrence, which is interpreted as an effective cloud thickness, is introduced. The underlying assumption in establishing the above relation is that cloud overlap approaches random overlap with increasing distance separating cloud layers and that the probability of deviating from random overlap decreases exponentially with distance. One month of Cloud‐Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) and CloudSat data (July 2006) support these assumptions, although the correlation length sometimes increases with separation distance when the cloud top height is large. The data also show that the correlation length depends on cloud top hight and the maximum occurs when the cloud top height is 8 to 10 km. The cloud correlation length is equivalent to the decorrelation distance introduced by Hogan and Illingworth (2000) when cloud fractions of both layers in a two‐cloud layer system are the same. The simple relationships derived in this study can be used to estimate the top‐of‐atmosphere irradiance difference caused by cloud fraction, uppermost cloud top, and cloud thickness vertical profile differences.

Orthogonal nonnegative matrix tri-factorization for co-clustering: Multiplicative updates on Stiefel manifolds

Matrix factorization-based methods become popular in dyadic data analysis, where a fundamental problem, for example, is to perform document clustering or co-clustering words and documents given a term-document matrix. Nonnegative matrix tri-factorization (NMTF) emerges as a promising tool for co-clustering, seeking a 3-factor decomposition X ≈ USV ⊤ with all factor matrices restricted to be nonnegative, i.e., U ⩾ 0 , S ⩾ 0 , V ⩾ 0 . In this paper we develop multiplicative updates for orthogonal NMTF where X ≈ USV ⊤ is pursued with orthogonality constraints, U ⊤ U = I , and V ⊤ V = I , exploiting true gradients on Stiefel manifolds. Experiments on various document data sets demonstrate that our method works well for document clustering and is useful in revealing polysemous words via co-clustering words and documents.

Insects associated with exposed decomposing bodies in the Colombian Andean Coffee Region

In Colombia, mainly classic forensic medicine methods were used to clarify crimes until 2004. However, other disciplines, including forensic entomology, started to be considered only after the New Accusatory System introduction in Bogotá and the Coffee Region in 2005. In order to provide tools for obtaining evidentiary material elements in judicial trials, it is presented here the succession of insects throughout the decomposition process of an exposed carcass of Sus scrofa Linnaeus 1758 (Suidae) and the Occurrence Matrix of colonizing species. This process was evaluated under ambient conditions in the Andean rural area of the city of Pereira, in the Mundo Nuevo district, located in a pre-montane Wet Forest area, from October to November 2006. A sampling period of 27 days and 3198 individuals were collected. We found these colonizing species in the following stages of decomposition: Lucilia eximia (Wiedemann, 1819) fresh; Hemilucilia semidiaphana (Rondani, 1850), Oxelytrum discicolle (Brullé, 1840), and Cochliomyia macellaria (Fabricius 1775) bloated; Chrysomya albiceps (Wiedemann 1819), Compsomyiops verena (Walker, 1849), Ophyra aenescens (Wiedemann, 1830) and Musca domestica Linnaeus, 1758 active; Fannia sp. advanced and Stearibia nigriceps (Meigen, 1826) remains. This study provides support tools to define the Post Mortem Interval that may be used by experts from government institutions and laboratories officially accredited.

The fish zonation of the Itanhaém river basin in the Atlantic Forest of southeast Brazil

The distribution range of fishes along an upstream–downstream gradient within a river basin is determined by the ecological requirements of each fish species. This differential pattern of distribution may suggest the occurrence of fish zones, where different functional guilds would prevail. The earliest concept of fish zonation highlighted the preferential position of several species in European rivers. Nowadays, there is a consensus that for a broad application of this concept, it is necessary to consider the division in functional groups along the downstream profile as related to reproductive, feed and population dynamics traits. However, for a number of tropical systems, such information does not exist. In this case, it is possible to use the fish family as a surrogate metric, by assuming that it fully covers common aspects of morphology, feeding, and behavior. Using this surrogate metric, we analyzed the distribution range of fish species along an elevation gradient in a typical Atlantic Forest basin in southeast Brazil. Our objective was to delimit fish zones based on the representativeness of some dominant fish families. Between 2006 and 2008, we sampled 42 sites from the estuary to the headwater creeks in elevations that varied from 3 to 783 m. The data were explored using the species (or family) occurrence, relative numerical abundance (N′), relative biomass (B′), and the importance index (IP). We applied the non-metric multidimensional scaling (NMDS) to ordinate the sampling sites using the Jaccard distance calculated on the species occurrence matrix. Data on family importance were superimposed to the NMDS to check how the distribution and the abundance of the families were correlated to the ordination. We caught 64 species distributed among 25 fish families. Based on the dominance of some families, we propose that the fish assemblage in the Itanhaem river basin can be organized in four fish zones (FZ) and three transitory zones (TZ), which replace each other along the longitudinal gradient. The marine/estuarine fishes dominated in FZ1 and TZ1. The freshwater fishes were dominated by Characidae (FZ2), Heptapteridae (FZ3), and Gymnotidae (FZ4). TZ2 was equally dominated by Characidae, Loricariidae, and Heptapteridae. The distinct assemblage composition along the FZ’s may relate to complex interactions involving seasonal patterns of variation in salinity levels (next to estuary), slope, presence of downstream barriers, and availability of food resources. We expect that the patterns observed in our study will also be presented in other coastal Atlantic Forest basins in Southeast Brazil.

Biogeography of Aphodiinae dung beetles based on the regional composition and distribution patterns of genera

AbstractAim To examine current biogeographical patterns of Aphodiinae dung beetles (Coleoptera: Scarabaeoidea: Scarabaeidae) in order to reveal relationships among regions and their potential impact on the diversification of this group.Location Worldwide.Methods Information about all Aphodiinae genera was obtained from the literature. An occurrence matrix was built for the six worldwide biogeographical regions, and their faunas were characterized through simple statistics. Regional variations and similarities were further explored using co‐occurrence and nestedness analyses, sequential agglomerative, hierarchical and nested clustering (SAHN), and a parsimony analysis of endemicity (PAE). Mantel tests were also employed to assess the relationships between several characteristics of the regions and their faunas.Results The Palaearctic and Palaeotropical regions showed the highest total numbers of Aphodiinae genera and the greatest generic endemism. Both these regions and the Oriental also showed higher numbers of genera than would be expected according to their size. Co‐occurrence and nestedness analyses confirmed the non‐randomness of the distribution of genera. Clustering and PAE showed that the Palaearctic and Oriental regions are the most similar, followed by the Palaeotropical region. Regional dissimilarity in genera composition was related to biological and historical traits, but not to ecoregions.Main conclusions A structured geographical pattern for Aphodiinae was confirmed. Land continuity and proximity in the long term could have played a unifying role in regional faunas. We suggest that the different biogeographical regions have acted as either macroevolutionary sources (basically the Palaearctic and the Palaeotropical regions) or sink regions, according to their role as diversification centres. We review the processes and events that could account for current patterns of Aphodiinae diversity.

New statistical approaches for anomaly detection

AbstractComputer security is a growing problem. Over the last years, the number and variety of security attacks in IP‐ based network infrastructures have increasingly grown, leading to the need of developing new security architectures. In this scenario, the use of intrusion detection systems (IDSs) has emerged as a key element, since it permits to tackle security threats by masquerader, misfeasor, and clandestine users. In this paper, we address the problem considering some new statistical techniques for detecting network anomalies. In more detail, the paper discusses the use of several statistical models to characterize the normal behavior of the network traffic running over TCP, so that anomalies can be revealed as significant deviations from such behavior. Namely, our proposal is based on the use of Markov chains, co‐ occurrence matrices, and compression algorithms, for modeling the TCP connections, in terms of statistical analysis of some of the packet header fields. The performance analysis, presented in this paper, demonstrates the effectiveness of the proposed methods. Copyright © 2009 John Wiley & Sons, Ltd.

The relation between Pearson's correlation coefficient r and Salton's cosine measure

AbstractThe relation between Pearson's correlation coefficient and Salton's cosine measure is revealed based on the different possible values of the division of the L1‐norm and the L2‐norm of a vector. These different values yield a sheaf of increasingly straight lines which together form a cloud of points, being the investigated relation. The theoretical results are tested against the author co‐citation relations among 24 informetricians for whom two matrices can be constructed, based on co‐citations: the asymmetric occurrence matrix and the symmetric co‐citation matrix. Both examples completely confirm the theoretical results. The results enable us to specify an algorithm that provides a threshold value for the cosine above which none of the corresponding Pearson correlations would be negative. Using this threshold value can be expected to optimize the visualization of the vector space.