Text-based Recommendation Systems for Software Developers: A Systematic Literature Review

Purpose: The aim of this SLR is to look at recommendation systems which receive textual information as an input. By analysing them it is possible to understand how the textual information is preprocessed and which algorithms are then used to generate recommendations. Methods: With the Search Query I frst identifed 487 papers, from which 65 were removed as duplicates. After the IC and EC application, 28 articles remained as relevant. Results: From these articles’ analysis, it was found that the most commonly used pre-processing techniques are tokenization, TF-IDF, and stopwords removal. I also determined that all algorithms for suggestions generation in such systems can be divided into 4 categories: classifcation, ranking, clustering, and heuristic-based algorithms. In the last step I found that the most frequent output of such systems are API, code, and workers suggestions. Conclusion: With this work, I looked at which pre-processing techniques are used in the text-based recommender systems for software developers and which are the most common. I have also looked at the classifcation of algorithms for such recommendation systems. Finally, I considered what kind of objects are recommended by these text-based recommendation systems.

Exploring Topic-language Preferences in Multilingual Swahili Information Retrieval in Tanzania

Habitual switching of languages is a common behaviour among polyglots when searching for information on the Web. Studies in information retrieval (IR) and multilingual information retrieval (MLIR) suggest that part of the reason for such regular switching of languages is the topic of search. Unlike survey-based studies, this study uses query and click-through logs. It exploits the querying and results selection behaviour of Swahili MLIR system users to explore how topic of search (query) is associated with language preferences—topic-language preferences. This article is based on a carefully controlled study using Swahili-speaking Web users in Tanzania who interacted with a guided multilingual search engine. From the statistical analysis of queries and click-through logs, it was revealed that language preferences may be associated with the topics of search. The results also suggest that language preferences are not static; they vary along the course of Web search from query to results selection. In most of the topics, users either had significantly no language preference or preferred to query in Kiswahili and changed their preference to either English or no preference for language when selecting/clicking on the results. The findings of this study might provide researchers with more insights in developing better MLIR systems that support certain types of users and in certain scenarios.

Prediction Candidates and Political Parties in the Presidential Election 2024 in Indonesia Based on Twitter

Twitter is a popular platform on social media that is used to predict presidential candidates and political parties who will contest in the presidential election. This study uses a quantitative approach with descriptive content analysis. This approach describes the details of a text or message related to discussions and information on the Twitter social network in the 2024 presidential election. The research subjects are Twitter social media users based on the involvement of Twitter users in the 2024 presidential election discourse in Indonesia. The data is obtained from Twitter with Twitter Search focusing on the keyword “Pilpres 2024”. The analytical tool used is Nvivo 12 Plus software with Word Frequency Query and Text Search Query analysis features. This study predicts that candidates with a strong chance as official candidates are Anies Baswedan, Prabowo, and Ganjar Pranowo. The mapping of political parties indicates that there will be political contestation between nationalist parties and religious-based parties in the 2024 presidential election.

Improving Search Engine Efficiency through Contextual Factor Selection

Learning to rank (LTR) is an important artificial intelligence (AI) approach supporting the operation of many search engines. In large-scale search systems, the ranking results are continually improved with the introduction of more factors to be considered by LTR. However, the more factors being considered, the more computation resources required, which in turn, results in increased system response latency. Therefore, removing redundant factors can significantly improve search engine efficiency. In this paper, we report on our experience incorporating our Contextual Factor Selection (CFS) deep reinforcement learning approach into the Taobao e-commerce platform to optimize the selection of factors based on the context of each search query to simultaneously maintaining search result quality while significantly reducing latency. Online deployment on Taobao.com demonstrated that CFS is able to reduce average search latency under everyday use scenarios by more than 40% compared to the previous approach with comparable search result quality. Under peak usage during the Single’s Day Shopping Festival (November 11th) in 2017, CFS reduced the average search latency by 20% compared to the previous approach.

Common communicable diseases in the general population in France during the COVID-19 pandemic

In France, social distancing measures have been adopted to contain the spread of COVID-19, culminating in national Lockdowns. The use of hand washing, hydro-alcoholic rubs and mask-wearing also increased over time. As these measures are likely to impact the transmission of many communicable diseases, we studied the changes in common infectious diseases incidence in France during the first year of COVID-19 circulation. We examined the weekly incidence of acute gastroenteritis, chickenpox, acute respiratory infections and bronchiolitis reported in general practitioner networks since January 2016. We obtained search engine query volume for French terms related to these diseases and sales data for relevant drugs over the same period. A periodic regression model was fit to disease incidence, drug sales and search query volume before the COVID-19 period and extrapolated afterwards. We compared the expected values with observations made in 2020. During the first lockdown period, incidence dropped by 67% for gastroenteritis, by 79% for bronchiolitis, by 49% for acute respiratory infection and 90% for chickenpox compared to the past years. Reductions with respect to the expected incidence reflected the strength of implemented measures. Incidence in children was impacted the most. Reduction in primary care consultations dropped during a short period at the beginning of the first lockdown period but remained more than 95% of the expected value afterwards. In primary care, the large decrease in reported gastroenteritis, chickenpox or bronchiolitis observed during the period where many barrier measures were implemented imply that the circulation of common viruses was reduced and informs on the overall effect of these measures.

Is Google Trends a reliable way to determine digital dental epidemiology?

Aim: Google Trends, which allows Internet users to interact with and search data, can provide in-depth information about new phenomena regarding population and health-related behavior and is a tool that can be accessed free of charge. With the widespread use of dental implants in almost every country in the world today, an increase has also been reported in the prevalence of peri-implantitis (PP), which is a peri-implant disease. The aim of this study is to determine whether the rates of PP that were obtained from previous studies on this disease are in line with the data obtained using Google Trends. Methodology: Using observational, ecological research, we searched Google Trends for the following query terms: peri implantitis + periimplantitis, to obtain the volume of this Internet search query. The queries were searched within Spain (ES), Germany (DE), the Netherlands (NL), the United Kingdom (UK), and Turkey from January 2010 to December 2019. Results: An examination of the search results for “peri-implantitis + peri-implantitis” on Google Trends found that the largest numbers of searches for these words were made from the country of ES, and the smallest numbers were made from Turkey. It took two years to make forecasts based on the results, and the study determined that there has been a change in the trends in countries that were searched for these words. Also, the results obtained in previous studies for the prevalence of peri-implantitis were not similar to the data obtained from Google Trends. Conclusion: We concluded in this study that Google Trends is not a reliable tool for dental epidemiology. How to cite this article: Üner DD, İzol BS. Is Google Trends a reliable way to determine digital dental epidemiology? Int Dent Res 2021;11(Suppl.1):38-46. https://doi.org/10.5577/intdentres.2021.vol11.suppl1.7 Linguistic Revision: The English in this manuscript has been checked by at least two professional editors, both native speakers of English.

The Multi-Domain International Search on Speech 2020 ALBAYZIN Evaluation: Overview, Systems, Results, Discussion and Post-Evaluation Analyses

The large amount of information stored in audio and video repositories makes search on speech (SoS) a challenging area that is continuously receiving much interest. Within SoS, spoken term detection (STD) aims to retrieve speech data given a text-based representation of a search query (which can include one or more words). On the other hand, query-by-example spoken term detection (QbE STD) aims to retrieve speech data given an acoustic representation of a search query. This is the first paper that presents an internationally open multi-domain evaluation for SoS in Spanish that includes both STD and QbE STD tasks. The evaluation was carefully designed so that several post-evaluation analyses of the main results could be carried out. The evaluation tasks aim to retrieve the speech files that contain the queries, providing their start and end times and a score that reflects how likely the detection within the given time intervals and speech file is. Three different speech databases in Spanish that comprise different domains were employed in the evaluation: the MAVIR database, which comprises a set of talks from workshops; the RTVE database, which includes broadcast news programs; and the SPARL20 database, which contains Spanish parliament sessions. We present the evaluation itself, the three databases, the evaluation metric, the systems submitted to the evaluation, the evaluation results and some detailed post-evaluation analyses based on specific query properties (in-vocabulary/out-of-vocabulary queries, single-word/multi-word queries and native/foreign queries). The most novel features of the submitted systems are a data augmentation technique for the STD task and an end-to-end system for the QbE STD task. The obtained results suggest that there is clearly room for improvement in the SoS task and that performance is highly sensitive to changes in the data domain.

Study protocol of the PIMPI-project, a cohort study on acceptance, tolerability and immunogenicity of second trimester maternal pertussis immunization in relation to term and preterm infants

Background Maternal immunization confers passive immunity to the fetus by transplacental antibody transfer. Infants may be better protected against pertussis if the mother received a diphtheriae, tetanus and acellular pertussis (Tdap) vaccination in the second trimester of pregnancy compared to the third trimester. This study evaluates IgG antibody concentrations in term and preterm infants at birth and 2 months after birth after maternal Tdap-vaccination between 200 and 240 w of gestation vs third trimester Tdap-vaccination. Further aims are assessing the determinants that underlie acceptance of second trimester maternal Tdap-vaccination as well as the tolerability of vaccination. Methods This prospective cohort study consists of two parts. In the acceptance part, pregnant women complete a questionnaire on determinants that underlie acceptance of a second trimester Tdap-vaccination, which is offered subsequently between 200 and 240 w of gestation. Vaccinated women complete an additional questionnaire on vaccination tolerability. Vaccinated women may also participate in the immunogenicity part, in which blood is drawn from mother at delivery and from infant at birth and 2 months after birth. Women are also eligible for the immunogenicity part if they received a Tdap-vaccination between 200 and 240 w of gestation via the national immunization program and get hospitalized for an imminent preterm delivery. Blood sampling continues until 60 term and 60 preterm mother-infant-pairs have been included. Pertussis-specific IgG antibody concentrations are determined in serum using a fluorescent bead-based multiplex immunoassay. For term infants, non-inferiority in IgG antibody concentrations against pertussis toxin (anti-PT) will be assessed referred to a historical control group in which mothers were Tdap-vaccinated between 300 and 320 w of gestation. For preterm infants, non-inferiority of anti-PT IgG concentrations is referred to as 85% of infants having ≥ 20 international units/mL at 2 months after birth. Discussion This study investigates acceptance, tolerability and immunogenicity regarding maternal Tdap-immunization between 200 and 240 w of gestation. Its results provide insight into the effects of second trimester Tdap-vaccination on IgG antibody concentrations in term and preterm infants before primary infant vaccinations. Results on acceptance and tolerability guide antenatal care providers in communication with pregnant women and maintain the safety of second trimester Tdap-vaccination. Trial registration: EU Clinical Trials Register, 2018-002976-41, retrospectively registered 24 July 2019, https://www.clinicaltrialsregister.eu/ctr-search/search?query=2018-002976-41.