Volume 35 - Special issue Data Intelligibilty, Business Intelligence and Semantic Web - 2022

1. A Hybrid Algorithm Based on Multi-colony Ant Optimization and Lin-Kernighan for solving the Traveling Salesman Problem

Mathurin Soh ; Baudoin Nguimeya Tsofack ; Clémentin Tayou Djamegni.
In this article, a hybrid heuristic algorithm is proposed to solve the Traveling Salesman Problem (TSP). This algorithm combines two main metaheuristics: optimization of multi-colony ant colonies (MACO) and Lin-Kernighan-Helsgaun (LKH). The proposed hybrid approach (MACO-LKH) is a so-called insertion and relay hybridization. It brings two major innovations: The first consists in replacing the static visibility function used in the MACO heuristic by the dynamic visibility function used in LKH. This has the consequence of avoiding long paths and favoring the choice of the shortest paths more quickly. Hence the term insertion hybridization. The second innovation consists in modifying the pheromone update strategy of MACO by that of the dynamic λ-opt mechanisms of LKH in order to optimize the solutions generated and save in execution time, hence the relay hybridization. The significance of the hybridization, is examined and validated on benchmark instances including small, medium, and large instance problems taken from the TSPlib library. The results are compared to four other state-of-the-art metaheuristic approaches. It results in that they are significantly outperformed by the proposed algorithm in terms of the quality of solutions obtained and execution time.

2. A Semantic Measure for Outlier Detection in Knowledge Graph

Bara Diop ; Cheikh Talibouya Diop ; Lamine Diop.
Nowadays, there is a growing interest in data mining and information retrieval applications from Knowledge Graphs (KG). However, the latter (KG) suffers from several data quality problems such as accuracy, completeness, and different kinds of errors. In DBpedia, there are several issues related to data quality. Among them, we focus on the following: several entities are in classes they do not belong to. For instance, the query to get all the entities of the class Person also returns group entities, whereas these should be in the class Group. We call such entities “outliers.” The discovery of such outliers is crucial for class learning and understanding. This paper proposes a new outlier detection method that finds these entities. We define a semantic measure that favors the real entities of the class (inliers) with positive values while penalizing outliers with negative values and improving it with the discovery of frequent and rare itemsets. Our measure outperforms FPOF (Frequent Pattern Outlier Factor) ones. Experiments show the efficiency of our approach.