Automated knowledge base management: A survey

Abstract

A fundamental challenge in the intersection of Artificial Intelligence and Databases consists of developing methods to automatically manage Knowledge Bases which can serve as a knowledge source for computer systems trying to replicate the decision-making ability of human experts. Despite of most of tasks involved in the building, exploitation and maintenance of KBs are far from being trivial, significant progress has been made during the last years. However, there are still a number of challenges that remain open. In fact, there are some issues to be addressed in order to empirically prove the technology for systems of this kind to be mature and reliable.

CoTO: A Novel Approach for Fuzzy Aggregation of Semantic Similarity Measures

“…The existence of many programming languages and individual coding styles poses a challenge for determining the degree of likeness between source code fragments [1]. Despite that, unsupervised similarity measures demonstrate reasonable adaptability when working with code written in multiple languages or according to various styles, so it can be accepted that they are good indicators of code similarity [28]. This feature is important in many scenarios, such as maintenance, whereby increased complexity is commonplace [15].…”

“…There are various techniques to assess the similarity between source code fragments, each with its unique perspective based on certain features of the code fragments being compared. The following is a list that we have already proposed previously [28], consisting of twenty-one distinct unsupervised approaches. Table 1 presents all these similarity measures.…”

“…According to existing research [28], it is unrealistic to think that these techniques alone can deliver performance equivalent to the state-of-the-art. However, we aim to show how their aggregation can do so under certain conditions.…”

“…• We build upon previous work [28] to curate a comprehensive collection of unsupervised similarity measures to compare the likeness of different code fragments. This collection facilitates a systematic exploration of the most effective measures for assessing source code similarity at low computational cost.…”

VCU at Semeval-2016 Task 14: Evaluating definitional-based similarity measure for semantic taxonomy enrichment

“…WordNet, Gene Ontology, and Medical Subject Headings. Building and maintaining these knowledge bases is expensive and manually intensive (Martinez-Gil, 2015). Semantic taxonomy enrichment aims to aid in the maintenance process by automatically placing new terms into an existing taxonomy.…”