Discovering contradicting protein-protein interactions in text

Bessis

2015

Soft Comput

Abstract-The extraction of meaningful, accurate, and relevant information is at the core of Big Data research. Furthermore, the ability to obtain an insight is essential in any decision making process, even though the diverse and complex nature of big datasets raises a multitude of challenges.In this paper, we propose a novel method to address the automated assessment of influence among concepts in big data sets. This is carried out by investigating their mutual cooccurrence, which is determined via topologically reducing the corresponding network. The main motivation is to provide a toolbox to classify and analyse influence properties, which can be used to investigate their dynamical and statistical behaviour, which would potentially lead to a better understanding and prediction of the properties of the system(s) they model.An evaluation was carried out on two real-world data-sets, which were analysed to test the capabilities of our system. The results show the potential of our approach, indicating both accuracy and efficiency.

Section: Discussion and Future Workmentioning

confidence: 99%

Section: Discussion and Future Workmentioning

confidence: 99%

An influence assessment method based on co-occurrence for topologically reduced big data sets

Bessis

2015

Soft Comput

“…In particular, specific disambiguation techniques (Manning CD, 1999) need to be implemented to enable a better and more general identification of the appropriate nodes in a BN by classifying and grouping together nodes referring to similar concepts. However, word synonymity and polysemy must be fully addressed, depending on the corresponding context (Sanchez-Graillet O, 2004), where a variety of supervised machine learning algorithms can be potentially utilised to facilitate this task. As discussed above, the investigation of suitable node attributes will also play an important role in the correct identification of the topology of the resulting networks.…”

Section: Text Analysis and Relation Extractionmentioning

confidence: 99%

“…BNs have proved to be very successful when a scenario consisting of preacknowledge information coupled with uncertain or partially known data, is considered (Pearl J, 1998). The extraction of BNs from text is typically a complex task due to the intrinsic ambiguity of natural language (Sanchez-Graillet O, 2004). In fact, BNs are defined by strict topological and probabilistic rules, which are difficult to fully automatise.…”

Section: Introductionmentioning

confidence: 99%

Automated extraction of fragments of Bayesian networks from textual sources

Hayes

Palmieri

et al. 2017

Applied Soft Computing

Mining large amounts of unstructured data for extracting meaningful, accurate, and actionable information, is at the core of a variety of research disciplines including computer science, mathematical and statistical modelling, as well as knowledge engineering. In particular, the ability to model complex scenarios based on unstructured datasets is an important step towards an integrated and accurate knowledge extraction approach. This would provide a significant insight in any decision making process driven by Big Data analysis activities. However, there are multiple challenges that need to be fully addressed in order to achieve this, especially when large and unstructured data sets are considered. In this article we propose and analyse a novel method to extract and build fragments of Bayesian networks (BNs) from unstructured large data sources. The results of our analysis show the potential of our approach, and highlight its accuracy and efficiency. More specifically, when compared with existing approaches, our method addresses specific challenges posed by the automated extraction of BNs with extensive applications to unstructured and highly dynamic data sources. The aim of this work is to advance the current state-of-the-art approaches to the automated extraction of BNs from unstructured datasets, which provide a versatile and powerful modelling framework to facilitate knowledge discovery in complex decision scenarios

“…Causal relationships are an example of the above which play an important role in a variety of knowledge discovery tasks with several applications [5]. More specifically, in [9], a method to automatically extract causal relationships to populate Bayesian Networks is introduced, also suggesting that for many applications it is important to consider influence rather than causality (see [7] for a detailed discussion).…”

Section: Introductionmentioning

confidence: 99%

Influence Discovery in Semantic Networks: An Initial Approach

2014 UKSim-AMSS 16th International Conference on Computer Modelling and Simulation

Bagdasar

2014

Abstract-Assessing the influence between concepts, which include people, physical objects, as well as theoretical ideas, plays a crucial role in understanding and discovering knowledge. Despite the huge amount of literature on knowledge discovery in semantic networks, there has been little attempt to fully classify and investigate the influence, which also includes causality, of a semantic entity on another one as dynamical entities. In this paper we will introduce an approach to discover and assess influence among nodes in a semantic network, with the aim to provide a tool to identify its type and direction. Even though this is still being developed, the preliminary evaluation shows promising and interesting results.