Joris Vertommen

A Personalized Tourist Trip Design Algorithm For Mobile Tourist Guides

Mobile tourist guides evolve towards automated personalized tour planning devices. The contribution of this article is to put forward a combined artificial intelligence and metaheuristic approach to solve tourist trip design problems (TTDP). The approach enables fast decision support for tourists on small footprint mobile devices. The orienteering problem, which originates in the operational research literature, is used as a starting point for modelling the TTDP. The problem involves a set of possible locations having a score and the objective is to maximize the total score of the visited locations, while keeping the total time (or distance) below the available time budget. The score of a location represents the interest of a tourist in that location. Scores are calculated using the vector space model, which is a well-known technique from the field of information retrieval. The TTDP is solved using a guided local search metaheuristic. In order to compare the performance of this approach with an algorithm that appeared in the literature, both are applied to a real data set from the city of Ghent. A collection of tourist points of interest with descriptions was indexed and subsequently queried with popular interests, which resulted in a test set of TTDPs. The approach presented in this article turns out to be faster and produces solutions of better quality.

Pairwise-adaptive dissimilarity measure for document clustering

This paper introduces a novel pairwise-adaptive dissimilarity measure for large high dimensional document datasets that improves the unsupervised clustering quality and speed compared to the original cosine dissimilarity measure. This measure dynamically selects a number of important features of the compared pair of document vectors. Two approaches for selecting the number of features in the application of the measure are discussed. The proposed feature selection process makes this dissimilarity measure especially applicable in large, high dimensional document collections. Its performance is validated on several test sets originating from standardized datasets. The dissimilarity measure is compared to the well-known cosine dissimilarity measure using the average F-measures of the hierarchical agglomerative clustering result. This new dissimilarity measure results in an improved clustering result obtained with a lower required computational time.

Multiple-vector user profiles in support of knowledge sharing

This paper describes an algorithm to automatically construct expertise profiles for company employees, based on documents authored and read by them. A profile consists of a series of high dimensional vectors, each describing an expertise domain, and provides a hierarchy between these vectors, enabling a structured view on an employees expertise. The algorithm is novel in providing this layered view, as well as in its high degree of automation and its generic approach ensuring applicability in an industrial setting. The profiles provide support for several knowledge management functionalities that are difficult or impossible to achieve using existing methods. This paper in particular presents the initialization of communities of practice, bringing together both experts and novices on a specific topic. An algorithm to automatically discover relationships between employees based on their profiles is described. These relationships can be used to initiate communities of practice. The algorithms are validated by means of a realistic dataset.

Topic identification based on document coherence and spectral analysis

Abstract In a world with vast information overload, well-optimized retrieval of relevant information has become increasingly important. Dividing large, multiple topic spanning documents into sets of coherent subdocuments facilitates the information retrieval process. This paper presents a novel technique to automatically subdivide a textual document into consistent components based on a coherence quantification function. This function is based on stem or term chains linking document entities, such as sentences or paragraphs, based on the reoccurrences of stems or terms. Applying this function on a document results in a coherence graph of the document linking its entities. Spectral graph partitioning techniques are used to divide this coherence graph into a number of subdocuments. A novel technique is introduced to obtain the most suitable number of subdocuments. These subdocuments are an aggregation of (not necessarily adjacent) entities. Performance tests are conducted in test environments based on standardized datasets to prove the algorithm’s capabilities. The relevance of these techniques for information retrieval and text mining is discussed.

A Quantified Approach to Tacit Knowledge Management in R&D-Environments through the Use of Document-Based User Profiles

Few corporate activities result in the density of knowledge generation that Research and Development activities do. This knowledge comes at high cost and should be managed as the high-value asset it is. Of paramount importance in the knowledge management process, is the avoidance of knowledge being unavailable for (re-)use within the enterprise. This paper discusses a quantified approach to the management of tacit knowledge through the creation of user profiles that capture the expertise of an employee. Once constructed, these profiles can be used to make tacit knowledge searchable, as well as to bring together experts and those in search of expertise.

Identifying Document Metadata Based on Multilayer Clustering

This paper presents a novel technique to semi-automatically identify metadata for documents when installing a knowledge management system. Document management systems often deal with large collections of documents. This vast amount of information needs to be searchable for the knowledge worker. Supporting techniques are needed to aid the knowledge worker in his search for information. Many of these techniques are based on the presence of metadata for each document. The techniques presented in this paper are based on a novel approach called multilayer clustering. Using this clustering technique, documents can be assigned to one or more document types. Besides this assignment to a specific type, properties and values are assigned to this document based on term networks extracted from this document. The preliminary tests presented in this paper were performed on a public and several private dataset. The results obtained from the tests indicate that this approach is promising.

Facilitating Product Development with the Help of Knowledge Management: the McKnow Platform

Product designers and other professionals spend a large part of their time in search of previously gained information and knowledge within and outside company borders. When unable to find what they are looking for, these professionals often resort to reinvention of the wheel. This paper introduces the McKnow platform, a framework on which algorithms are developed that support a quantified approach to knowledge management. Key aspects of these algorithms are automation and user-orientation. Several offered functionalities are described, as well as conclusions from their experimental evaluation.