October 1, 2025 Thesis Open Access

Hayu Bekele

MARC21 XML Export

<?xml version='1.0' encoding='UTF-8'?>
<record xmlns="http://www.loc.gov/MARC21/slim">
  <leader>00000nam##2200000uu#4500</leader>
  <datafield tag="653" ind1=" " ind2=" ">
    <subfield code="a">Afaan Oromo Spelling Checker and Corrector, Deep Learning Spelling Checker,  LSTMSpell, GRUSpell, Artificial Injection</subfield>
  </datafield>
  <controlfield tag="005">20251001062544.0</controlfield>
  <controlfield tag="001">19291</controlfield>
  <datafield tag="856" ind1="4" ind2=" ">
    <subfield code="s">2409625</subfield>
    <subfield code="z">md5:4b76bbe34b12455c29aecfd03a4115d0</subfield>
    <subfield code="u">https://zenodo.org/record/19291/files/Hayu Bakele.pdf</subfield>
  </datafield>
  <datafield tag="542" ind1=" " ind2=" ">
    <subfield code="l">open</subfield>
  </datafield>
  <datafield tag="260" ind1=" " ind2=" ">
    <subfield code="c">2025-10-01</subfield>
  </datafield>
  <datafield tag="909" ind1="C" ind2="O">
    <subfield code="p">user-ddu-business-economics</subfield>
    <subfield code="o">oai:zenodo.org:19291</subfield>
  </datafield>
  <datafield tag="100" ind1=" " ind2=" ">
    <subfield code="a">Hayu Bekele</subfield>
  </datafield>
  <datafield tag="245" ind1=" " ind2=" ">
    <subfield code="a">Design and Develop Spelling Checker and  Corrector for Afaan Oromoo Using Deep Learning  Approach</subfield>
  </datafield>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">user-ddu-business-economics</subfield>
  </datafield>
  <datafield tag="540" ind1=" " ind2=" ">
    <subfield code="u">http://www.opendefinition.org/licenses/cc-by</subfield>
    <subfield code="a">Creative Commons Attribution</subfield>
  </datafield>
  <datafield tag="650" ind1="1" ind2="7">
    <subfield code="a">cc-by</subfield>
    <subfield code="2">opendefinition.org</subfield>
  </datafield>
  <datafield tag="520" ind1=" " ind2=" ">
    <subfield code="a">&lt;p&gt;In this digital era, texts are provided to many text editing and processing tools for copious tasks. Search engines, emailing and word processing are top word entry and editing envi ronments in different languages where the faultlessness is a keynote for the intent of that text being processed. For the support of this goal, many languages used a spelling checker and corrector in their text processing tools. A spelling checker is an application, program or simply a software that arbitrate the correctness of the spelling for a given word based on the languages&amp;rsquo; spelling rule. With the fact of its meritorious, different languages are being used spelling checkers in their document processing tasks. Afaan Oromoo, spoken by the largest ethnolinguistic group, constituting more than one-third of Ethiopian population, is one of federal official language now a days. To its digital advancement, Afaan Oromo needs such a computer based support of text editing and processing tool. Hence, in this thesis we have designed and developed spelling checker and corrector for Afaan Oromoo using deep learning algorithms. We gathered and prepared dataset of 37311 Afaan Oromo sentences which consists 609272 words. These data are collected from media news of Afaan OromoprogramFacebookpages like Oromia Broadcasting Corporate (OBN), Oromia Media Network (OMN), Fana Broad casting Corporate (FBC) and Voice of America (VoA) scraped using python script. In addi tion to thesesources, wealsocollectedAfaanOromocorpusfromseveralpolitical, academic, scientific, and fiction books and as well as new testament Afaan Oromo holy bible book. These data finally, preprocessed and spited in to train and test with 80:20 ratio respectively. In this thesis work, we used deep learning algorithms Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) models of Recurrent Neural Network (RNN) with their bidirectionalities. We developed four model using LSTM, GRU, BiLSTM, BiGRU models for Afaan Oromo spelling checker and corrector. Precision, recall, F-scoreand, Support evalu ation metrics of scikit learn are used to compare their performance with each other on the same test data. Finally, we preferred BiGRU model with 92.325%, 83.5% and 91.008% of precision, recall and F1 score respectively&lt;/p&gt;</subfield>
  </datafield>
  <datafield tag="773" ind1=" " ind2=" ">
    <subfield code="n">doi</subfield>
    <subfield code="i">isVersionOf</subfield>
    <subfield code="a">10.20372/nadre:19290</subfield>
  </datafield>
  <datafield tag="024" ind1=" " ind2=" ">
    <subfield code="a">10.20372/nadre:19291</subfield>
    <subfield code="2">doi</subfield>
  </datafield>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">publication</subfield>
    <subfield code="b">thesis</subfield>
  </datafield>
</record>