scholarly journals Hash Tables for a Digital Lexicon

2021 ◽  
Vol 28 (2) ◽  
pp. 25-38
Author(s):  
Fábio Carlos Moreno ◽  
Cinthyan Sachs C. de Barbosa ◽  
Edio Roberto Manfio

This paper deals with the construction of digital lexicons within the scope of Natural Language Processing. Data Structures called Hash Tables have demonstrated to generate good results for Natural Language Interface for Databases and have data dispersion, response speed and programming simplicity as main features. The storage of the desired information is done by associating a key through the hashing functions that is responsible for distributing the information in this table. The objective of this paper is to present the tool called Visual TaHs that uses a sparse table to a real lexicon (Lexicon of Herbs), improving performance results of several implemented hash functions. Such structure has achieved satisfactory results in terms of speed and storage when compared to conventional databases and can work in various media, such as desktop, Web and mobile.

2020 ◽  
Vol 2 (1) ◽  
pp. 22-31
Author(s):  
Dewi Soyusiawaty ◽  
Anna Hendri Soleliza Jones

Daftar judul skripsi yang ada di web digilib.uad.ac.id belum digunakan secara optimal. Kesulitan menentukan topik serta pengelolaan data pada sistem yang belum memadai menjadi beberapa kendala mahasiswa. Penelitian ini bertujuan membangun pencarian informasi skripsi dengan antarmuka bahasa alami agar mudah dalam menulis kriteria pencarian tanpa harus terikat formulir pencarian. Penelitian ini menggunakan data skripsi untuk dikelola. Pengambilan data dilakukan menggunakan pendekatan Natural Language Processing. Masukan dalam bentuk kalimat bahasa alami digunakan untuk mencari data.  Proses Parsing dilakukan untuk memecah kalimat input dan mendeteksi kata kunci yang relevan. Pengembangan aturan produksi dalam Context Free Grammar diperlukan untuk menerjemahkan bahasa alami ke dalam query.  Kalimat yang melewati tahap parser diterjemahkan ke dalam bahasa SQL. Sistem ini berhasil menampilkan informasi skripsi berupa daftar judul berdasarkan topik, metode, dan objek penelitian sesuai kalimat pencarian dengan nilai precision sebesar 89,3% dan recall sebesar 100%. Keberadaan model pencarian informasi dengan antarmuka bahasa alami dapat menjadi alternatif dalam proses pencarian informasi skripsi guna menyediakan sistem yang lebih fleksibel.  The use of the existing digital library has not been used optimally. Difficulties in determining topics and managing data in an inadequate system are among the obstacles for students. This study aims to build a thesis information search with a natural language interface so that it is easy to write search criteria without having to be tied to a search form. This research uses thesis data to be managed. Data were collected using the Natural Language Processing approach. Input in the form of natural language sentences is used to find data. The parsing process is carried out to break down the input sentence and detect relevant keywords. Development of production rules in Context Free Grammar is necessary to translate natural language into queries. Sentences that go through the parser stage are translated into SQL language. This system succeeds in displaying thesis information in the form of a list of titles based on topics, methods, and research objects according to the search sentence with a precision value of 89% and a recall of 100%. The existence of an information retrieval model with a natural language interface can be an alternative in the thesis information search process in order to provide a more flexible system.


2020 ◽  
pp. 3-17
Author(s):  
Peter Nabende

Natural Language Processing for under-resourced languages is now a mainstream research area. However, there are limited studies on Natural Language Processing applications for many indigenous East African languages. As a contribution to covering the current gap of knowledge, this paper focuses on evaluating the application of well-established machine translation methods for one heavily under-resourced indigenous East African language called Lumasaaba. Specifically, we review the most common machine translation methods in the context of Lumasaaba including both rule-based and data-driven methods. Then we apply a state of the art data-driven machine translation method to learn models for automating translation between Lumasaaba and English using a very limited data set of parallel sentences. Automatic evaluation results show that a transformer-based Neural Machine Translation model architecture leads to consistently better BLEU scores than the recurrent neural network-based models. Moreover, the automatically generated translations can be comprehended to a reasonable extent and are usually associated with the source language input.


Diabetes ◽  
2019 ◽  
Vol 68 (Supplement 1) ◽  
pp. 1243-P
Author(s):  
JIANMIN WU ◽  
FRITHA J. MORRISON ◽  
ZHENXIANG ZHAO ◽  
XUANYAO HE ◽  
MARIA SHUBINA ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document