Search

Fast technology development in recent years has led to a rapid increase in data generation. Most of the data are raw and need to be classified and processed into clean, usable data. However, classifying a huge volume of data is not an easy task, and it requires massive human efforts to label the data. To assist humans and shorten this labelling process time, data clustering is one of the most used methods in the first step of data cleaning. Among popular data clustering algorithms, Hierarchical Clustering is one of the best that can help to separate noise to recognize data patterns. In this paper, we conducted intensive experiments to verify the effectiveness of Hierarchical Clustering on multiple datasets using Apache Spark. We evaluated Hierarchical Clustering by comparing its performance in different sizes. Their volumes are provided in the experiment section.

This paper presents a method for building a real estate chatbot automatically to support customers in Vietnamese. The chatbot is trained with data set collected on Facebook groups and from the famous real estate website in Vietnam. Using Logistic Regression, user’s intent recognition task achieves precision = 0.93, recall = 0.87 and F1-score = 0.89, while the automatic entity labeling achieves 83% accuracy thanks to the development of a real estate knowledge base. Besides, we report our experience on the design of dialog management modules.

The number of smart phone and smartphone users has increased rapidly, boosting the growth of mobile advertising. Predicting the percentage of users who click on an advertisement when displayed on an application (CTR: Click Through Rate) plays an important role in choosing an advertising push strategy for each application. However, because the field of mobile advertising is still new, the study of solutions to these problems is only for Web advertising. This paper presents a new predictive method based on Matrix Factorization (MF) algorithm. The paper also proposes a method of integrating the reliability of data into MF algorithms. The method was tested on actual data provided by Amobi advertising company with 1.7 million records including 300 advertisements and 400 mobile device applications. The AUC (Area Under The Curve) value obtained on the test set is 0.9037 with MF and 0.9411 when integrating reliability into the MF algorithm. Thus, the results show that the predictive quality is significantly improved when integrating data reliability and evaluation into MF algorithms.

DBpedia is a community project upon extracting information from Wikipedia and standardize this data with Semantic Web and Linked Data principles. Therefore, it is a large-scale knowledge base and considered as one of the currently emerging heated issues in the subject of Semantic Web. Moreover, thanks to the development of DBpedia community, many more opportunities are created in the Linked Open Data (LOD) area. These lead to giving birth to the specific DBPedia chapters in different languages. However, there are many languages that do have not their own chapters including Vietnamese. The reason is that the extraction procedure of DBpedia requires manual mappings which match Wikipedia templates to corresponding types and properties of DBpedia ontology. Thus, in this paper, we describe how to identify automatically types for Vietnamese entities as the first step toward building the Vietnamese DBpedia chapter.

This paper focuses on describing the process of building a linked open data repository for presenting tourist resources, which presently contains more than 100,000 places in big cities of Vietnam. This information is extracted and integrated from various sources on the Internet, which are famous social networks or tourism websites, for example, Booking.com, Mytour, Vinalo, etc. These massive and heterogeneous materials would be described via a semantic knowledge base presented in RDF/XML format that is accessible by SPARQL query. Concentrating in studying the concept of linked open data with its publishing processes, this work adjusts and improves the repository to become a standardized linked open data and allows it connect with other famous ontology such as DBpedia.

Purpose – The purpose of this paper is to propose an automatic method to generate semantic annotations of football transfer in the news. The current automatic news integration systems on the Web are constantly faced with the challenge of diversity, heterogeneity of sources. The approaches for information representation and storage based on syntax have some certain limitations in news searching, sorting, organizing and linking it appropriately. The models of semantic representation are promising to be the key to solving these problems. Design/methodology/approach – The approach of the author leverages Semantic Web technologies to improve the performance of detection of hidden annotations in the news. The paper proposes an automatic method to generate semantic annotations based on named entity recognition and rule-based information extraction. The authors have built a domain ontology and knowledge base integrated with the knowledge and information management (KIM) platform to implement the former task (name...

Part 1: Semantic Concepts and Open Data; International audience; News on the Internet today plays an important role in helping people access daily information around the world. News aggregators are websites that collect and provide content from different sources in one location for easy viewing. However, the increasing number of news on the Internet makes it difficult for readers when they desire to access news they are concerned. One solution to this issue is based on employing recommender systems. In this research, we propose a novel method for news recommendation based on a combination of semantic similarity with content similarity between news and implement it as a feature of semantic-based news aggregators BKSport. Experimental results have shown that, a combination of both kind of similarity measures will result in better recommendation than when using either measure separately.

In education organizations, a number of quotidian and duplicated consulting questions for training programs, university regulations... make difficulties, inconvenient and time-consuming for consultants. This work proposes a method to suggest answers of similar questions, which may help both students and consultants in orientation. This method includes two main tasks: (i) query generation which extracts key phrases from input questions, (ii) searching which find out most similar questions with their answers and relevant scores using extracted key phrases. For the first task, a check strategy was used to build key phrase candidates and the Naive Bayes Classifier was used to select key phrases. The two different approaches were experimented for the second one: (i) Similarity Comparison Searching, and (ii) Sorl Search Engine. The precision of the key phrase extraction is about 69 %. BKCA, an e-consultancy system for SoICT-HUST, was built and experimented with about 46 % relevant question-answer pairs for the Similarity Comparison searching approach, and about 39 % for the Sorl Search Engine one.

Nowadays, Linked Open Data (LOD) has grown rapidly to become large open datasets defined by RDF standards. Thanks to development of Data Web, the information on LOD is increasingly deeper, larger and easier to link in multi-domains, constituting the Linked Open Data cloud. Recently end-users applications using linked data sources as background knowledge appeared [15]. Thus, the exploitation of information on LOD effectively brings tremendous values as well as challenges. Meanwhile, Exploratory Search (ES) describes information-seeking processes that are opportunistic, iterative, and multi-tactical [7]. Furthermore, systems based on ES capitalize on new technological capabilities and interface paradigms that facilitate an increased level of interaction with information. In this paper, we present a method on searching and recommending information to empower exploratory search with semantic technologies. Our aim is to use algorithms with incorporation of structured semantics in search to give users the best related-semantic results and enhance users' interactions. We have exploited the data within LinkedMDB1 to support users in finding some information in the movie domain.

In recent years, the botnet phenomenon is one of the most dangerous threat to Internet security, which supports a wide range of criminal activities, including distributed denial of service (DDoS) attacks, click fraud, phishing, malware distribution, spam emails, etc. An increasing number of botnets use Domain Generation Algorithms (DGAs) to avoid detection and exclusion by the traditional methods. By dynamically and frequently generating a large number of random domain names for candidate command and control (C&C) server, botnet can be still survive even when a C&C server domain is identified and taken down. This paper presents a novel method to detect DGA botnets using Collaborative Filtering and Density-Based Clustering. We propose a combination of clustering and classification algorithm that relies on the similarity in characteristic distribution of domain names to remove noise and group similar domains. Collaborative Filtering (CF) technique is applied to find out bots in each botnet, help finding out offline malwares infected-machine. We implemented our prototype system, carried out the analysis of a huge amount of DNS traffic log of Viettel Group and obtain positive results.