School of Information Science

Permanent URI for this college

http://etd.aau.edu.et/handle/123456789/26

Browse

Browsing School of Information Science by Author "Abdela Woinshet (W/ro)"

Now showing 1 - 5 of 5
  • No Thumbnail Available
    Item
    The Application of Websom for Amharic Text Retrieval
    (Addis Ababa University, 2003-06) Mamuye Bizuneh; Amsalu Saba (W/rt); Biru Tesfaye (Ato); Abdela Woinshet (W/ro)
    This research explored the applicability of WEBSOM (Web Based Self Organizing map) for retrieving texts written in Amharic language. The method applies a neural network's self organizing algorithm for generating the map display. The map display detects complex relationships among given documents, and reveals the relationships based on the arrangements of terms abstracted from the documents. To conduct the experiment, 330 Amharic news articles of three classes were collected from the Ethiopian News Agency. 248 of the news articles were taken as a training set and the remaining as a test set. For the purpose of document representation, the Vector Space Model was used. Non-content bearing terms were removed from the lists of terms identified from the headline and slug parts of the news articles and suffix/prefix-stripping technique was applied on the remaining list. After changing terms having different writing forms in to one common form, terms with a total frequency of above 70 and below 3 were discarded from the list. Then, a matrix both for the training and test set were constructed on the remaining 142 terms. A normalized weight was assigned to each term in a given news article based on TF-IDF (Term Frequency- Inverse Document Frequency) weighting technique and the vector matrix were prepared in appropriate format for the tool to be used. Using Nenet (Neural Network Tool), the SOM map was trained with the 248 articles in the training set and tested with three test sets selected from the three classes of news articles. From the distribution of these articles on the map, it was observed that the map placed similar articles near to each other. The results obtained from the three tests made, indicated that the clustering capability of the SOM for Amharic documents is promising. x Lastly, a map was constructed for the entire (330) news articles and an HTML based prototype browsing interface map was developed and labled with descriptive terms that convey properties of the area. A link was also made with the actual database through the Active Server Pages created so that users can browse on the map for relevant articles.
  • No Thumbnail Available
    Item
    Automatic Categorization of Amharic News Text: a Machine Learning Approach
    (2003-07) Teklu Surafel; Bekele Rahel (W/ro); Abdela Woinshet (W/ro); Lamnew Workshet (Ato)
    Currently newspaper companies and news agencies in Ethiopia are implementing a manual categorization system to categorize Amharic news articles in their day-to-day activities (although they are using computer system to store and dispatch information). The objective of this research was to investigate the application of machine learning techniques to automatic categorization of Amharic news items. 11, 024 news articles were used to do this research. To come up with good results text preparation and preprocessing was done. Stop-word and words that occur in 3 or less documents were removed from the collection. Thirty-three percent of the data was used for testing purposes. Machine learning techniques, Naïve Bayes and k Nearest Neigbor classifiers, were used to categorize the Amharic news items. The result of this research indicated that such classifiers are applicable to automatically classify Amharic news items. However, the classifiers work well when the categories contain almost evenly distributed news items. The best result obtained by the naïve Bayes and kNN classifiers is on three categories data (95.80% vs. 89.61%) and the least performance is shown on the 16 categories (78.48% vs. 64.50%) respectively. The 16 categories contain unevenly distributed data than the three categories and it is learnt that unevenly distributed numbers of documents over the categories decreases the performance of both classifiers; K nearest Neighbor dramatically decreases than naïve Bayes. This research indicated that Naïve Bayes is more applicable to automatic categorization of Amharic news items. The result of this research is promising. Nevertheless, additional works are recommended in order to come up with good result. Keywords: Text categorization, machine Learning, naïve Bayes, K Nearest Neigbor
  • No Thumbnail Available
    Item
    Design and Development of Automatic Morphological Synthesizer for Amharic Perfective Verb Forms
    (Addis Ababa University, 2002-06) Lisanu Kibur; Admasu Yonas (PhD); Abdela Woinshet (W/ro); Berhanu Solomon
    Natural Language processing plays a significant role in increasing computers’ capability to understand natural languages. Morphological synthesis is one aspect of the task of understanding natural language, the language by which most human knowledge is recorded. Morphological synthesis or generation is a process of returning one or more surface forms from a sequence of underlying (lexical) forms. Morphological synthesis systems are used as components in many applications, including machine translation, spell-check, speech recognition, dictionary (lexicon) compilation, POS tagging, morphological analysis, conversational systems, automatic sentence construction and many others. Today, synthesizers of different kinds have been developed for languages that have relatively wider use internationally. The same cannot be said for Amharic, the working language of the Federal Government of Ethiopia, and one of the major languages of the country (Bender, 1976). This study is, thus, an attempt to develop a prototype automatic morphological synthesizer for Amharic, specifically for perfective verb forms. In this study, algorithms that take into consideration the morphological properties of Amharic are developed from scratch and applied, as there are no previous such attempts. The study adopts the combination of rule-based and neural network approaches to design and develop a prototype, referred as AmharicMorphologicalSynthesizer. The rule-based approach generates all the roots successfully where as the neural network predicts the type of roots in the test data set with an accuracy of 81.48%. For a new case consisting of 14 roots the neural network identifies type A perfective verb forms with an accuracy of 80%, type B perfective verb forms with accuracy of 25% and that of type C perfective verb forms with an accuracy of 100%. The thesis, in short, describes processes of automated morphological synthesis from manually synthesizing words to developing a prototype and conducting an experiment with it. The result obtained using the small manually constructed root table will encourage the undertaking of further research in the area, especially with the aim of developing a full-fledged Amharic morphological synthesizer.
  • No Thumbnail Available
    Item
    Incorporation of Multimedia Features in Conventional Cataloguing Database: a Case Study of the Institute of Ethiopian Studies Museum
    (Addis Ababa University, 1997-05) Abdela Woinshet (W/ro); Biru Tesfaye (PhD)
    Researchers and curators at museums are increasingly faced with providing meticulous inventories of the objects under their care. They are confronted with a dearth of information about relationships between form, function and use of objects, as well as information about production methods, use and users. The IES museum is not an exception to this. Among the major activities preoccupying the museum is information provision on its holdings: maintenance of accession registers, card catalogues, photograph documentation, treatment records etc. This study attempted to address the information handling problems existing 111 the IES museum. Currently manual methods of information management are in place. These methods, however, have created problems in information handling activities because of the size and diversity of the collection as well as varied information requirements of the users. Previous studies suggested conventional computer-based bibliographic database systems as one means of dealing with such problems. Although the recommendations may work well for IES in general and as a starting point for the museum in particular, the user survey conducted in this regard showed that the recommendations are of limited Use to enhance the information handling activity of the museum as records in the museum objects comprise different data types, i.e., structured and Unstructured text, still images, sound, and video, and which are difficult to fit into the conventional formats. To this end, an attempt is made in this work to demonstrate the incorporation of multimedia features in the suggested cataloguing databases in order to enhance systematization in the organization and accessibility of the museum objects at the Institute.
  • No Thumbnail Available
    Item
    A Survey of Information Requirements of Agricultural Experts: The Case of Ministry of Agriculture in Kenya
    (Addis Ababa University, 2001-06) Osoro Leah; Jemaneh Getachew (PhD); Abdela Woinshet (W/ro)
    This study surveys the information requirements of agricultural experts in MoA, Kenya. Specifically, to assess the information needs; to find out the extent to which these needs are met; to establish from where the experts obtain information; to assess the information seeking behavior of agriculturists; to find out the problems encountered by the experts when seeking information and to offer suggestions towards effective information delivery to the agriculturists based on the survey results.Agriculture information plays a very important function in agricultural management. Regardless of this fact, the currently existing agricultural information systems within Ministry of Agriculture do not fully perform their vital role in agricultural management, as it is believed to. In particular, the Library Information Services (LIS) has failed to provide information to its users; the library is disseminating outdated information, relied on infrequently revised monographs at the expense of current sources of information (Kampuchea, 1998). The available information is ineffectively disseminated; therefore, there is a need to improve linkage between agricultural information and agricultural management, in view of the fact that the current paradigm is to set up agricultural information systems that tie the users to the information resources.A survey, usmg a semi-structured questionnaire was used to gather data. Seventy-six agricultural experts were sampled using stratified and purpose sampling techniques. The data was analyzed by use of descriptive statistics The study revealed that the experts' information needs were greatly influenced by the nature of the work they do, and the information needs in tum influences their information seeking habits. The agricultural experts have placed high importance on teclmical, market, and socioeconomic information. They rely on informal sources of information such as colleagues. They do not rely on library collections to satisfy their information needs. It was noted that card catalogue as an information retrieval tool is still dominant in the MoA libraries. Limiting factors to effective use of agricultural information by the experts was mainly due to the fact that the information sources are scattered at different locations. To address the problem of information sources scattered at different places within MoA and outside sources, a Ready Reference database system for MoA was proposed, designed and a prototype developed. The objective of the proposed database system is to direct users to sources of information that satisfy their information needs. It was also recommended that: Selective dissemination of information should be a major component of agricultural information services, teaching of information seeking skills, formulation of an agricultural information policy, management support, modernization of the MoA information services, future studies to consider applying other data-gathering techniques like interviewing, observing the businesses of MoA in operation and examining documents in particular those used to record or display information among others.