Browsing by Author "Alemayehu, Nega (PhD)"
Now showing 1 - 9 of 9
Results Per Page
Sort Options
Item Design and Development of Amharic Word Parser(Addis Ababa University, 2000-05) Bayou, Abiyot; Alemayehu, Nega (PhD)Most of the human knowledge is recorded using natural language. This language is too complex to be understood by computers. Only computers that have the capability to understand Natural language can access all these information efficiently. in addition, Natural Language interfaces to computers would allow complex systems to he accessible to everyone. Activities in the area of Natural Language Process ing tries to design and implement computer programs that can understand Natu ral I"anguagc and act appropriate ly on the infol111ation contained in the text or utterance. Enabl ing computers to understand natural language involves extraction of mea ning from words in texts. Word parsing is a process of extracting the meaning expressed by the 0rds of a text by to kenis in words to their basic components. This study is an attempt o f des igning and implementing word parser for Amhari c words. The study focuses on Amhari c verbs illldthe ir derivations, which covers more th an 75% of the words in the language. The paper describes the morpho logy of Amharic verbs and attempts to extract features that enables to implement automattc Amhari c word parser. The morphological data described and analyses in this study is taken from study outputs in the area of Amharic word morphology. An algorithm that can use the morphological property of the language has been designed and presented. Even tho ugh this studly focuses on Amharic verbs and their derivations the algorithm and the discuss ion can be easily ext ended I(lr other word groups. The study also describes the information contained in affixes and designs a database that can hold this information for word parsing and further analysis works.Item Development of a Stemming Algorithm for Afaan Oromoo Text(Addis Ababa University, 2000-06) Mekonnen, Wakshum; Alemayehu, Nega (PhD)This paper reports the design and development of a stemming algorithm for Afaan 01"011100 language. Reviews of Afaan Oromoo morphology, stemming algorithms, and other relevant materials were made. In Afaan Oromoo, inflectional and derivational affixations are the major word formation processes. The initial design of the stemming algorithm was based on free-context conflation procedures following the longest-match suffix removal approach. An accuracy rate of 71 % was obtained from this initial attempt. The improved algorithm incorporated suffix, context-sensitive, and recording reles in the procedures. Before stemming, functional and frequently occurring words, which were compiled as stop list, are excluded from the input term(s) to increase the efficiency of the stemmer. Procedures for prefix removal and for conflation of words formed by reduplication of first syllable are also components of the modified algorithm . Using the modified stemmer an accuracy rate of92% was gained from the test based on a sample of 1061 words. The percentage of errors recorded as understenuning and over stemming were reduced to 4.58% and 2.5% respectively from 10.5% and 17.5% for the first version. A substantial decrease in size of sample text is achieved from this stemmer. The morphological complexity of the language is the main sources of errors for the resulting inaccuracies of the stemming algorithm. For further improvement of the stemmer therefore, detailed study ofaan oromoo morphology is helpful. The result of this study in general shows the possibility of employing a stemming algorithm for conflating Afaan orpmpp words.Item Feed Analysis Information Retrieval System for Ruminant Livestock in Sub-Saharan Africa(Addis Ababa University, 1996-06) Atieno, Rachel; Chowdhury, G.G. (PhD); Alemayehu, Nega (PhD)Feed information is a valuable resource for animal production users which should be made available in the appropriate format to meet the user need, This study examined the need for an automated Feed Analysis Information Retrieval System for Sub-Saharan Africa, as a means for improving the overall animal production in the region. To attain the objectives of the study, the existing feed information systems were reviewed and analyzed to identify factors which limit these systems' availability and use . A common phenomena identified was lack of standard formats for documentation and characterization of the animal feed stuffs in the region.Additionally, most of the information on the composition of the indigenous African feedstuffs were mainly found at Research Institutions where their availability was limited to researchers of such institutions. Surveys undertaken as pan of this work revealed that this information is important to individuals and groups involved in animal production and should be made widely available in the region; many institutions and individuals are currently depending on foreign Feed Tables which may not always be appropriate for the situation in Sub-Saharan Africa. To enhance the use of, and accessibility to this information. a prototype feed analysis information retrieval system (FAIRS) has been developed using d Base IV. FAIRS has two main components: the user interface and the Database. The Database structure has been designed using a relational data model. In the data structure, International Feed Standards and codes have been incorporated to facilitate standardization of the system, and information exchange at institutional and regional levels. The second component viz the interfaces has been designed to facilitate interactive system operations by both the experienced and novice users. The system's function and user interfaces are designed with simple user instructions at all levels.It is recommended that available feed composition information be compiled and incorporated in this or similar databases , which can then be produced on CD-ROM and made available to National Agricultural Research Systems. Internet connectivity is increasingly becoming available to countries in Africa, a development which should be taken advantage. The possibility for providing feed information on the Internet should be examined by regional and international institutions like lLRl, with Information Technology capabilities.Item A Hypertext Based Computer Aided Instruction System in Amharic: A Case Study with the Course Introduction to Computers.(Addis Ababa University, 1996-05) Dessalegn, Girma; Chowdhury, G.G (PhD); Alemayehu, Nega (PhD)The application of computers to the process of teaching and learning holds many prognoses. Their use has been seen as a way of improving the quality of delivery, increase the individualization of learning, encouraging learning privacy, provide greater motivation, improve information retention. Other advantages include immediate feedback, the opportunity of leasable students to benefit from repeated use, and more active participation by students. of all the possible applications of computer technology to learning, hypertext seems to be one ""ith a great potential Hypertext allows learners to access randomly stored information based on their needs, interest, or whims. It focuses learner attention on the relationship of facts. It also encourages active student -centered learning since there are always decisions about what to read next. Learning in one's own language is believed to be important. One of the most important advantages of learning in one's own language is that learning begins with something that is already known The need for training on the basics of computers is strongly felt in Ethiopia. This high demand of producing computer littermate people is being responded by the mushrooming of computer training institutions in Addis Ababa and the major cities of the country. It was identified that the training situation, being in its infancy, is faced with many problems. This study attempts to demonstrate the design and development of a hypertext based courseware in Amharic by considering a case study on the course 'Introduction to Computers' which is being offered at the National Computer and Information Centre of Ethiopian Science and Technology Commission. It also tries to address problems regarding the computer training situation in the country. A literature review on the definition, advantage, design consideration, types, and current trends and developments on hypertext and courseware is presented. Different models of hypertext and courseware development are discussed. The models are used to identify the steps and processes involved in the development process of a hypertext courseware which satisfies instructional System principles. The quick prototyping software system development approach is used to develop the course ware. Knowledge Pro for Windows is used to develop the program.Item Information Support System for Export and Import in Ethiopia(Addis Ababa University, 1995-05) Sebsbie, Gezahegne; Bhattacharyya, G. (Prof.); Alemayehu, Nega (PhD)This study has been carried out to design and develop an information support system for export and import in Ethiopia. It is to provide different items of information needed by the different user groups as per their requirement. The design of the system has been preceded by a survey conducted to collect information about the following: (I) The prospective users of export and import Information; (2) the different types of information needed by these groups of users; (3) the purposes for which these items of information are required; (4) the different sources of export and import information; and (5) the existing available trade information service units and the types of information service rendered by these units. The provisions of the Ethiopian National Information Policy have been taken note of. The trade information infrastructure of the country has been reviewed in the light of the provisions of the national information policy to recognize its inadequacy. For the purpose of collecting information, methods, such as, questionnaire method, interview method, observation method, and literature survey, have been used. The study reveals that the existing trade information service units are neither adequately resourceful; nor are they well organized. In general, the performance of these trade information service units is not at all satisfactory according to their users. Those service units are all under-utilized because of their inadequate in all respects. Being in conformity with the expectations of the national information policy, a computer based information services system for export and import in Ethiopia has been proposed. The system is to be located in the information service unit of the Ministry of Trade, to start with. In the long run the system would function as a network. All the important trade information service units of the country would function as nodes of this network. For the proposed system to be operative, it would be necessary for certain essential development to take place mostly through the necessary initiatives of the Government of Ethiopia. The recommendation of this study cover all those essential development.Item Online Tutorial for Micro CDS/ISIS Software(Addis Ababa University, 1995-05) Fisseha, Sisay; Alemayehu, Nega (PhD); Chowdhury, G.G. (PhD)User assistance through online help or tutorial facilities are among the major considerations in designing a user interface which in turn is the major factor for successful utilization of a software. Micro CDS/ISIS is an information retrieval software package developed by UNESCO and distributed free of cost to non-profit making organization. in spite of its free availability and large number of users, use of the software has been severely limited by a number of factors, chief among them is lack of adequate user assistance and instruction through online help or tutorial facilities . This thesis is an attempt to design and develop an online tutorial facility for Micro CDS/ISIS software which is believed to contribute greatly to ease of use and learning the software. In accomplishing this task, the software has been analyzed and described in terms of its major functions. Survey of the available developmental approaches has also been conducted and the one appropriate for the current work has been adopted. The tutorial consists of twelve lessons. It is developed using Turbo Vision which is an object oriented application framework for windowing programs. Although all the lessons have been designed , due to time constraints only the first three lessons have been programmed. However, the major design features proposed for the tutorial have been incorporated in these three lessons. Among the features that the tutorial provides is use of menus to provide smooth access to the tutorial lessons , and provision of control button that enables the learner to control the pace of the tutorial , linking of the concepts thereby enabling the learner to get more information on related concepts , use of animated windows and highlights in drawing attention and motivation of learner etc .Item Online Tutorial for Micro Cdsiisis Software(Addis Ababa University, 1995-05) Fissaha, Sisay; Alemayehu, Nega (PhD)User assistance through online help or tutorial fac ilities are among the major considerations in des igning a user interface which in turn is the major factor for successful utilization of a software. Micro CDS/ISIS is an infomlation retrieval software package developed by UNESCO and distributed free of cost to· non-profit making organization. In spite of its free availabiliry and large number of users , use of the software has been severely limited by a number of factors , --------cchieramonp hemis-la-cIc-oradequate-user-assistalIce-and-insrructionLhrough-orrline-help-or---i tutorial facilities. This thesis is an attempt to design and develop an online tutorial faciliry for Micro CDS/ISIS software which is believed to contribute greatly to ease of use and learning the software. In accomplishing this task, the software has been analyzed and described in terms of its major functions. Survey of the available developmental approaches has also been conducted and the one appropriate for the current work has been adopted. The tutorial consists of twelve lessons . It is developed using Turbo Vision which is an object oriented application framework for windowing programs. Although all the lessons have been designed, due to time constraints only the first three lessons have been programmed. However, the major design features proposed for the tutorial have been incorporated in these three lessons. Among the features that the tutorial provides is use of menus to provide smooth access ro the tutorial lessons , and provision of control button that enables the learner to control the pace of the tutorial , linking of the concepts thereby enabling the learner to get more information on related concepts, use of animated windows and highlights in drawing attention and moti vation of learner etc .Item Word Sense Disambiguation for Amharic Text Retrieval: A Case Study for Legal Documents(Addis Ababa University, 2009-04) Kassie, Teshome; Alemayehu, Nega (PhD)This study demonstrates how linguistic d disambiguation based on semantic vector analysis can improve the effectiveness of an Amharic document query retrieval algorithm. Accurate document retrieval based on query criteria is important in every knowledge do main. The ability to retrieve appropriate documents is made more difficult by the fact that many words can have different meanings in different contexts. If search engines could disambiguate those words, more accurate retrieval of documents should be able to be achieved. For this study, an Amharic disambiguation algorithm was developed based on the principles of semantic vectors and implemented in Java. The disambiguation algorithm was then used to develop a document search engine. A set of 865 Ethiopian Amharic language legal statute documents were selected as the document population that would be searched. Ten queries containing Amharic keywords with ambiguous meaning were selected. An expert was used to identify which documents should ideally be retrieved by each query. Depending on the query, the expert identified between 6 and 25 documents that should be retrieved. The semantic vector query algorithm created in this study was compared to the well known Lucienne algorithm. Each query was run using both algorithms. The 20 most relevant documents were identified for each query from each algorithm . For each query, the list of documents retrieved by each algorithm was compared to the list of documents identified by the expert. The number of correct (consistent with the expert's choices) documents retrieved by each algorithm was measured. Results are that the semantic vector algorithm was superior for 6 of the 10 queries (Lucene was superior o n 2 queries, and o n two they were tied). This difference was not statistically significant. However, if the total number of correct document id gentrification are taken into account (not just which algorithm was superior for each query) the n the semantic vector algorithm averaged 82% correct identification of documents w here as the Lucienne algorithm was only 49% accurate. This difference was highly statistical y significant (p <0.02) less t han the level of significant (pItem Word Sense Disambiguation for Amharic Text Retrieval: A Case Study for Legal Documents(Addis Ababa University, 2009-04) Kassie, Teshome; Alemayehu, Nega (PhD)This study demonstrates how linguistic d disambiguation based on semantic vector analysis can improve the effectiveness of an Amharic document query retrieval algorithm. Accurate document retrieval based on query criteria is important in every knowledge do main. The ability to retrieve appropriate documents is made more difficult by the fact that many words can have different meanings in different contexts. If search engines could disambiguate those words, more accurate retrieval of documents should be able to be achieved. For this study, an Amharic disambiguation algorithm was developed based on the principles of semantic vectors and implemented in Java. The disambiguation algorithm was then used to develop a document search engine. A set of 865 Ethiopian Amharic language legal statute documents were selected as the document population that would be searched. Ten queries containing Amharic keywords with ambiguous meaning were selected. An expert was used to identify which documents should ideally be retrieved by each query. Depending on the query, the expert identified between 6 and 25 documents that should be retrieved. The semantic vector query algorithm created in this study was compared to the well known Lucienne algorithm. Each query was run using both algorithms. The 20 most relevant documents were identified for each query from each algorithm . For each query, the list of documents retrieved by each algorithm was compared to the list of documents identified by the expert. The number of correct (consistent with the expert's choices) documents retrieved by each algorithm was measured. Results are that the semantic vector algorithm was superior for 6 of the 10 queries (Lucene was superior o n 2 queries, and o n two they were tied). This difference was not statistically significant. However, if the total number of correct document id gentrification are taken into account (not just which algorithm was superior for each query) the n the semantic vector algorithm averaged 82% correct identification of documents w here as the Lucienne algorithm was only 49% accurate. This difference was highly statistical y significant (p <0.02) less t han the level of significant (p