Showing papers by "Elizabeth Sherly published in 2014"
01 Dec 2014
TL;DR: A rule based approach for sentiment analysis from Malayalam movie reviews is proposed, which gives the polarity at the sentence level for the movie reviews with an accuracy of 85%, when analysed.
Abstract: This paper proposes a rule based approach for sentiment analysis from Malayalam movie reviews. The research in Sentiment Analysis nowadays become one among active research areas in natural language processing. Sentiment Analysis is the cognitive process in which the user's feeling and emotions are extracted. The growing importance of sentiment analysis coincides with the growth of social media such as reviews, forum discussions, blogs, and social networks. Sentiment analysis enables computers to automate the activities performed by human for making decisions based on the sentiment of opinions, which has wide applications in data mining, web mining, and text mining. Negation Rule has been applied for extracting the Sentiments from a given text. This system gives the polarity at the sentence level for the movie reviews with an accuracy of 85%, when analysed.
29 citations
01 Dec 2014
TL;DR: The paper presents an algorithm to identify the pronominals and its antecedents in the Malayalam text input by employing a hybrid of statistical machine learning and rule based approaches.
Abstract: Anaphora resolution (AR) is the process of resolving references to an entity in the discourse The paper presents an algorithm to identify the pronominals and its antecedents in the Malayalam text input Anaphora resolution is achieved by employing a hybrid of statistical machine learning and rule based approaches The system is implemented by exploiting the morphological richness of the language and it makes use of parts of speech tagging, subject-object identification and person-number-gender of the NPs We outline a simple, efficient but a naive algorithm for anaphora resolution, which computes the salience value score for each antecedents The system performance is evaluated with precision, recall measures which produced promising results The anaphora resolution system itself can improve the performance of many NLP applications such as text summarisation, text categorisation and term extraction
4 citations
01 Dec 2014
TL;DR: This paper concentrates on refactoring XMI of Sequence diagram, an XML Meta data Interchange, with OCL constraints to build a framework for automatic code generation and the proposed model is tested in a coal mill of a Thermal Power Plant, a highly complex time constrained system.
Abstract: The UML Sequence Diagram along with Model Driven Architecture in software development helps to model the time constraint behavior that enhances the legibility of the structure and behavior of a system. The Object Constraint Language (OCL) helps to convey additional constraints and invariants required, but OCL confines into an expression language. The lack of program logic and flow of control limit these models to generate codes and also for proper verifications. This paper concentrates on refactoring XMI of Sequence diagram, an XML Meta data Interchange, with OCL constraints to build a framework for automatic code generation. The proposed model is tested in a coal mill of a Thermal Power Plant, a highly complex time constrained system. The source code generated from the refactored XMI is able to generate the set of coal mill parameters that matches to the real plant data results.
4 citations
TL;DR: An algorithm for alignment-free sequence comparison using Logical Match, which compute the score using fuzzy membership values which generate automatically from the number of matches and mismatches is proposed.
Abstract: This paper proposes an algorithm for alignment-free sequence comparison using Logical Match. Here, we compute the score using fuzzy membership values which generate automatically from the number of matches and mismatches. We demonstrate the method with both the artificial and real datum. The results show the uniqueness of the proposed method by analyzing DNA sequences taken from NCBI databank with a novel computational time.
2 citations
03 Apr 2014
TL;DR: The top five algorithms such as Logistic, Bagging, LMT, Multiclass classifier and Attribute selection classifier which can be used for image classification are highlighted and identified.
Abstract: Identifying the wide range of applications, machine learning algorithms proved its ability to learn without being explicitly programmed. Classifying the images through machine learning algorithms is getting wide range of acceptability nowadays. Being a branch of Artificial Intelligence, machine learning implies the study of systems which has the capability to learn from data. Machine learning involves two parts - representation and generalization. Representation implies labeling seen data instances and generalization determines whether the system can perform well on unlabelled data instances. In this article, we focused on the performance of machine learning algorithms [1]. A CBIR (Content Based Image Retrieval) frame work has been developed and obtained a reduced texture feature data set using Caltech101 image database [2]. We highlight the top five algorithms such as Logistic, Bagging, LMT, Multiclass classifier and Attribute selection classifier which can be used for image classification. In introduction, an overview of the selected techniques is presented. We have extracted 2037 feature vectors from Caltech101 image database. These data are used to distinguish the performance of machine learning algorithms. Having checked all machine learning algorithms supported, we identified top five algorithms that have a better performance compared to other machine learning algorithms. The software used for testing is WEKA [3], which is an open source software developed by University of Waikato, New Zealand.
2 citations
TL;DR: A deconverting generator for Malayalam language using Universal Networking Language (UNL) for Machine Translation, which involves identifying the dependent features like syntactic, semantic and lexical features of target language.
Abstract: This paper presents a deconverting generator for Malayalam language using Universal Networking Language (UNL) for Machine Translation. UNL being an Interlingua representation, conveyed as directed hyper graph with relations and attributes of source language sentence. A set of Universal Words are generated from the source language with its semantic representation, are mapped to UNL features. The work involves identifying the dependent features like syntactic, semantic and lexical features of target language. UNL Relations, UNL Attributes and Universal Word (UW), which are the building blocks of UNL are identified and mapped to the dependent features of Malayalam. Lexical mapping of UWs to root words of Malayalam was done through UNL-Malayalam Word Dictionary. The deconversion is tested against 100 Malayalam Sentences that has achieved an appreciable F-measure score of 0.978. . General Terms Malayalam Deconversion, Universal Networking Language, Interlingua Machine Translation.
1 citations
TL;DR: In this article, the Logical Match algorithm was used to mine the indices of the sequential pattern in the correlation matrix memory (CMM) for sequential data mining, and the uniqueness of the method was demonstrated with both the artificial and the real datum taken from NCBI databank.
Abstract: This paper proposes a method for sequential data mining using correlation matrix memory. Here, we use the concept of the Logical Match to mine the indices of the sequential pattern. We demonstrate the uniqueness of the method with both the artificial and the real datum taken from NCBI databank.
1 citations
TL;DR: This paper is in its novel attempt to incorporate the notion of HITS algorithm with the utilities of Ontologies for the effective web document clustering, which could in tremendous way enhance the present information retrieval engulfed in the hurdles of accessing, extracting, interpreting and finding relevant information as is expected by the user themselves.
Abstract: The challenging aspects of immensely huge information in WWW poses a huge threat of retrieving correct information at the correct instant of time. When world is growing in a fast pace so is the exponential growth of information available in millions forms [4]. So retrieving information matching in synonym and in the right context is unmanageable in the current scenario. The need of the hour is to establish a system that could intelligently decipher the context and inherently extract the correct implicit meaning as what is expected by the user searching for information. Main contribution of this paper is in its novel attempt to incorporate the notion of HITS algorithm with the utilities of Ontologies for the effective web document clustering, which could in tremendous way enhance the present information retrieval engulfed in the hurdles of accessing, extracting, interpreting and finding relevant information as is expected by the user themselves.
Journal Article•
TL;DR: A new approach for automatic registration of mammograms and MR images that gives fast and accurate registration that can be effectively used for detecting masses and micro-calcifications with high level of accuracy is presented.
Abstract: The mammogram and MRI used for detection of cancerous cells in breast, requires overlaying the approach to get the best out of two or a combined effort. However, the varying intensity of breast in both modalities make the automatic segmentation of such lesions extremely challenging as both the modalities are different. This paper presents a new approach for automatic registration of mammograms and MR images. The registration of images taken from MRI and mammography is carried out to locate the position of lesion present in breast followed by noise removal, edge enhancement and morphological operations. The chest wall removal is done to extract the breast area, for obtaining the mass area and micro-calcifications to locate its position. The locations of lesions in both images are extracted out using region growing algorithm and thresholding method. Registering the images provide the most accurate location of masses and micro-calcification that is missed in any of the modalities, into single image. The results show that the method gives fast and accurate registration that can be effectively used for detecting masses and micro-calcifications with high level of accuracy. The accuracy obtained in the proposed method is about 87%. Keywords—image pre-processing, chest wall removal, nipple localization, segmentation, registration.