Masnizah Mohd

Design of an Interface for Interactive Topic Detection and Tracking

This paper presents the design of a new interface for interactive Topic Detection and Tracking (TDT) called Ievent . It is composed of 3 main views; a Cluster View, a Document View, and a Named Entity View, supporting the user in identifying new events and tracking them in a news stream. The interface has also been designed to test the usefulness in interactive TDT of named entity recognition. We report some initial findings from a user study on the effectiveness of our novel interface.

Enhanced Arabic Information Retrieval: Light Stemming and Stop Words

Stemming is a process of reducing inflected words to their stem, base or root from a generally written word form. For languages that is high inflected like Arabic. Stemming improve the retrieval performance by reducing words variants. The effectiveness of stop words lists with light stemming for Arabic information retrieval (General stopwords list, Khoja stopwords list, Combined stopwords list), were investigated in this paper. Using vector space model as the popular weighting scheme was examined. The idea is to combine (General and Khoja) stopwords lists with light stemming to enhance the performance, and compare their effects on retrieval. The Linguistic Data Consortium (LDC) Arabic Newswire data set was used. The best performance was achieved with the Combined stopwords list, with light stemming.

Effect of ISRI stemming on similarity measure for arabic document clustering

Arabic Document Clustering has increasingly become an important task for obtaining good results with the unsupervised learning task. This paper aims to evaluate the impact of the five measures (Cosine similarity, Jaccard coefficient, Pearson correlation, Euclidean distance and Averaged Kullback- Leibler Divergence) for Document Clustering with two types of pre-processing morphology-based The Information Science Research Institute (ISRI) is equivalent to the root-based stemmer and light stemmer; and without stemming without morphology) for an Arabic dataset. Stemming is known as a computational process used to reduce words to their stems. For classification, it is categorised as a recall-enhancing or precision-enhancing component. It is concluded that the method of ISRI for words is proved to be better than without stemming methods which use a five similarities/distance measures for Document Clustering.

A Study on the Naturalness of Gesture-Based Interaction for Children

The emergence of new gesture-technologies that use bare-hands without any remote control or tools to hold is a good indicator that the technology is implementing naturalness in an input control system. However, the concept of naturalness that is commonly applied is interpreted from the adult users perspective without realizing that the equally importance users of gesture-based technology are children. For this reason, this study was undertaken to describe the natural elements of gesture-based interaction in terms of how they influence the behavior of children using gesture-based technology devices, and to what extent the children benefit from their use. This includes the identification of issues and opportunities related to naturalness in using the latest gesture-based technologies, Kinect and iPad. Our observations show that the naturalness in gesture-control devices enabled children to reflect real world situations into the interaction, thus, aiding them to call back (recall and demonstrate) the gesture command easily and sparking positive feelings during the interaction. We conclude that understanding naturalness from a childrens perspective can offer potential benefits to children in the utilization of gesture-based technologies.

Evaluation of an interactive topic detection and tracking interface

Interactive Topic Detection and Tracking (iTDT) refers to the TDT works which focus on user interaction, user evaluation and user interfaces aspects. This article investigates and identifies elements of the design of an interface that aims to facilitate journalists performing TDT tasks such as tracking and detection. It presents an (iTDT) interface called Interactive Event Tracking (iEvent), and evaluates the usability of the features introduced. The findings indicate the features that facilitated the participants in performing both tasks: cluster labelling and top terms features in Cluster View, a histogram with the timeline and document content features in Document View, and a keyword approach feature in Term View. Meanwhile, features such as cluster visualisation in Cluster View and histogram with the timeline in Term View only facilitated participants during the tracking task. The study shows that the interface enables journalists to perform well in TDT tasks.

i-JEN: visual interactive Malaysia crime news retrieval system

Supporting crime news investigation involves a mechanism to help monitor the current and past status of criminal events. We believe this could be well facilitated by focusing on the user interfaces and the event crime model aspects. In this paper we discuss on a development of Visual Interactive Malaysia Crime News Retrieval System (i-JEN) and describe the approach, user studies and planned, the system architecture and future plan. Our main objectives are to construct crime-based event; investigate the use of crimebased event in improving the classification and clustering; develop an interactive crime news retrieval system; visualize crime news in an effective and interactive way; integrate them into a usable and robust system and evaluate the usability and system performance. The system will serve as a news monitoring system which aims to automatically organize, retrieve and present the crime news in such a way as to support an effective monitoring, searching, and browsing for the target users groups of general public, news analysts and policemen or crime investigators. The study will contribute to the better understanding of the crime data consumption in the Malaysian context as well as the developed system with the visualisation features to address crime data and the eventual goal of combating the crimes.

Construction of topics and clusters in Topic Detection and Tracking tasks

This paper discussed the construction of topics to be tracked and clusters to be detected in Topic Detection and Tracking (TDT) tasks. Single Pass Clustering was used to cluster the news articles. As a result, the TDT tasks contained a combination of a good and poor clustering performance based on the F1-measure. Therefore, the selection of clusters and topics from the clustering experiment is important in the Tracking and the Detection tasks. It has contributed towards the user experimental design.

A comparison of named entity patterns from a user analysis and a system analysis

This paper investigates the detection of named entity (NE) patterns by comparing the results of NE patterns resulting from a user analysis and a system analysis. Findings revealed that there are difference in NE patterns detected by system and user, something that may affect the performance of a TDT system based on NE detection.

COMPARATIVE STUDY OF K-MEANS AND K-MEANS++ CLUSTERING ALGORITHMS ON CRIME DOMAIN

This study presents the results of an experimental study of two document clustering techniques which are k-means and k-means++. In particular, we compare the two main approaches in crime document clustering. The drawback of k-means is that the user needs to define the centroid point. This becomes more critical when dealing with document clustering because each center point represented by a word and the calculation of distance between words is not a trivial task. To overcome this problem, a k-means++ was introduced in order to find a good initial center point. Since k-means++ has not being applied before in crime document clustering, this study presented a comparative study between k-means and k-means++ to investigate whether the initialization process in k-means++ does help to get a better results than k-means. We proposes the k-means++ clustering algorithm, to identify best seed for initial cluster centers in clustering crime document. The aim of this study is to conduct a comparative study of two main clustering algorithms, namely k-means and k-means++. The method of this study includes a pre-processing phase, which in turn involves tokeniza-tion, stop-words removal and stemming. In addition, we evaluate the impact of the two similarity/distance measures (Cosine similarity and Jaccard coefficient) on the results of the two clustering algorithms. Exper-imental results on several settings of the crime data set showed that by identifying the best seed for initial cluster centers, k-mean++ can significantly (with the significance interval at 95%) work better than k-means. These results demonstrate the accuracy of k-mean++ clustering algorithm in clustering crime doc-uments.

Personal information management systems and interfaces: An overview

In this paper we present an overview of a Personal Information Management (PIM) System and some examples of the interface applications. PIM involves methods and procedures to store, manage, retrieve and show information. PIM systems are becoming more ubiquitous and present a need for their functions toward improving and enhancing information search results. Therefore, studying the current PIM systems becomes unavoidable in order to develop newer systems. The purpose of this review is to discover and describe the criteria of user interfaces in PIM systems. We first describe the state of the art in PIM systems and present reviews on the elements of PIM user interfaces. Then we discuss the interface components between applications and illustrate the features provided in each.