Journal of Intelligent Manufacturing

, Volume 27, Issue 1, pp 83–99 | Cite as

The ALTER-NATIVA knowledge management approach

  • Joao Sarraipa
  • Catarina Marques-Lucena
  • Silvia Baldiris
  • Ramón Fabregat
  • Silvana Aciar
Article

Abstract

Nowadays, it is commonly known that information systems need an agile capability of handling knowledge. To accomplish this, systems have to have a formal knowledge representation ability supported by specific and advanced reasoning features. This research work proposes a knowledge management approach with the purpose to gather, model and consume community knowledge for specific recommendation commitments. Such approach is accomplished by a semantic lexicon alignment between the various community knowledge assets, to facilitate collaborations establishment between people and systems in an interoperable fashion. Thus, a knowledge base supported by a thesaurus able to represent all the metadata needed to represent and characterize the various community stakeholders’ resources is proposed. The thesaurus represents the lexicon in the domain, which in the ALTER-NATIVA systems is mostly used to support the various e-Learning elements (e.g. courses) and users categorization, sustained by synchronization features to facilitate a constant update of its information. A set of services designed to recommend specific resources in relation to a determined profile of user is provided. Additionally, a discussion about how the ALTER-NATIVA knowledge management approach can be applied to industrial environments is presented.

Keywords

Knowledge management E-learning Ontologies 

Introduction

A true e-business of the twenty-first century should ensure that employees have at their fingertips the knowledge, applications and services needed for the job (May 2002). Nowadays, information systems that commonly provide such knowledge, intends to deliver it to the right person with the required context at the right time, right place, in the right form.

Then, the practice of sharing such knowledge reveals and expands our individual and corporate competencies, which once understood can be reinforced and redistributed via e-Learning (Rhoads 2002). In that case, a practical e-Learning development approach can facilitate and promote the development of competencies and knowledge in industry (Sarraipa et al. 2013). It follows the idea that training is one of the basic means of human resources development in business organizations, aiming to motivate employees, to develop their potential and to help them perform better (Roy and Raymond 2008). Many companies adopt strategies to maintain employees up to date and consequently encourage continuing education (learning) (Rolstadås 2011). Consequently having knowledge management and e-Learning working together, it is a powerful source due to its easiness providing users their requirements in the form of correct and complete information (Chawhan 2012).

Currently is observed a great need for the development or generation of services or mechanisms of adaptive and intelligent e-Learning content. E-learning is commonly referred to the intentional use of networked information and communication technology in teaching and learning. The main objective is to create a common platform for integration of knowledge and content management with collaboration techniques (Bhattacharya et al. 2010), to enhance knowledge transfer and performance in education and training. These technologies offer learners control over content, learning sequence, pace of learning, time, and often media, allowing them to tailor their experiences to meet their personal learning objectives (Ruiz et al. 2006).

Virtual education services should be accessible to everyone (teachers and students). For that reason is necessary to consider the specific needs of each user, and consequently adapt the process to solve those needs in a dynamic way. This can be done through the formal representation of knowledge that can be used to support adaptive e-Learning services, even when it is addressed diversity characteristics.

Diversity is here addressed due to its relevance in today’s policy, but also because industries would like to be better introduced in communities even acting as inclusion catalysts. Ignoring diversity issues costs time, money, and efficiency. Some of the consequences can include unhealthy tensions; loss of productivity because of increased conflict; inability to attract and retain talented people of all kinds; complaints and legal actions; and inability to retain valuable employees, resulting in lost investments in recruitment and training (UCSF 2012).

Research problem

Knowledge is considered the key asset of modern organizations and industry. Thus, academia has been working to provide the appropriate support to leverage this asset. Some examples are: the extensive work on knowledge models and knowledge management tools (Costa et al. 2013). However, it still needs to have appropriate knowledge management approaches able to handle semantic interoperability issues mainly between different organisation and communities. This means that creating and leveraging individual and group knowledge in terms of organizational knowledge for practical usage and allocation still remains as an unsolved problem (Chu et al. 2010). Therefore, it is necessary that the information must be understood, commented, organized and transformed for problem solving (Karayel et al. 2004). That’s why knowledge modelling is one of the key activities in understanding, designing, implementing, and operating systems. Thus, to understand a domain, it is necessary a good model of that domain knowledge (Camarinha-Matos and Afsarmanesh 2007). Authors with the presented work contribute for this issue resolution.

Authors start by presenting a knowledge base establishment process that intends to describe how can an interoperable community information model be defined. Then, it is presented the various ALTER-NATIVA knowledge base elements followed by its aggregated platform. It integrates various modular interfaces, which through specific synchronization functionalities are able to constantly present its data updated in its different components, facilitating at the same time various channels for knowledge maintenance procedures. Finally, a set of advanced services based on the knowledge management features is presented. Such services intend to demonstrate the added value of using such knowledge management approach. At the end some conclusions and future work is also presented.

A knowledge base establishment process

When an information system intends to represent a domain knowledge needs to be aligned to the community that it represents. Consequently it is required to have a solution where community members could present their view on the domain and discuss it with their peers. Additionally, such knowledge must be available and maintained by all the involved actors. To facilitate this it is needed to have a knowledge acquisition and modelling process, which intends to establish the reference knowledge on the domain. The process chosen by authors to be followed it is based on MENTOR—methodology for enterprise reference ontology development (Sarraipa et al. 2010).

The reference knowledge is defined following all the steps of the MENTOR light version that are: (1) terminology gathering; (2) glossary building; (3) thesaurus building; (4) ontology building (Sarraipa 2012).These steps have as outputs the following elements respectively: terminology; glossary; thesaurus; and ontology. After having this ontology populated with instances representation of the real world becomes a reference Knowledge Base (KB) on the domain. Due to their specific purpose representing in a formal way the knowledge of a specific domain, these outputs are so-called Knowledge Representation Elements (KREs). These KREs are shown in Fig. 1. They are positioned in a specific order illustrating the taxis path to build knowledge of a specific domain (Sarraipa 2012). “Taxis” comes from the Greek and means arrangement and to place in order (thefreedictionary.com 2003), which in this specific case represents a specific arrangement, structured, of the lexicon to build knowledge.
Fig. 1

KREs relations in the taxis path to build knowledge

Terminology is the first KRE of the proposed taxis build of knowledge. It is composed by words that are commonly used to describe its domain contexts. Such terminology is the input for the glossary. A glossary is a set of terms and their definitions, and is bound to the domain where these definitions are set, which can be used when communicating information in order to unify knowledge sharing. It is similar to a domain dictionary. The difference is that a dictionary can have specific relations related to a concept (e.g. synonyms; translations).

The glossary content is then used as input to the thesaurus definition on the domain. A thesaurus is a structure that manages the complexities of terminology and provides conceptual relationships, ideally through an embedded classification of the type “is a”. The idea of having a thesaurus is related to the easiness to which, further on, could be used as a good input to the ontology structure definition and for human concepts searching process. Thesaurus and ontologies hierarchical structures are usually referred as taxonomies. Taxonomy is a classification system that categorizes all the information in a class/subclass relationship, representing a simple tree structure. At the top of this structure is a single classification, the root node that applies to all objects. The root node represents most general category of all things that the domain is related to. Nodes below this root are more specific classifications that apply to subsets of the total set of classified objects (Wikipedia.org 2013). For any category, each subcategory is a new taxonomy.

Consequently, a thesaurus is like taxonomy of domain concepts composed by its reference meanings. Ontology uses the thesaurus as input and accomplishes it with the enrichment of its concepts, with properties and rules and interrelations between them, with the purpose of following the view of a determined group, represent a segment of the reality (domain). At this stage, the ontology is an unspecified conceptual system, which underlies a particular KB (Guarino and Giaretta 1995). By other words, ontologies provide a structure for development of KBs, serving to define its models (O’Leary 1998). When ontologies are associated with real instances/individuals they become KBs.

In information systems, two main parts usually compose a KB about a specific domain: (1) the meta-knowledge, and (2) the domain lexicon. Meta-knowledge includes information about the knowledge that the system possesses, like the characteristics of the methods or plans used by the system, etc. The meta-knowledge is generally used to guide future planning or execution phases of the system (University of Michigan EECS Department 2013). The domain lexicon is the vocabulary related to a particular subject, or a list of concepts necessary to express a specific domain, which is understandably shared by a group of individuals.

Figure 2 illustrates the described knowledge management approach for an information system establishment. The domain lexicon is defined by domain experts, which provide concepts and definitions (semantics) to the building of a thesaurus in the domain. Knowledge engineers define the meta-knowledge. The result is an ontology, which would be prepared to integrate (i.e. through merging) the thesaurus defined by the domain experts. Such ontology would constitute information about the system and other data characteristics, which is usually represented in standards, about the domain. These mentioned characteristics represent the “syntax” of the domain vocabulary.
Fig. 2

A knowledge management approach

Two distinct branches that derive into a KB compose the proposed knowledge management approach. This KB supports all the domain knowledge and facilitates the knowledge integration between all the actors of the system. These branches aggregate automatic synchronizations in such way that if there is a new concept added by a person to the lexicon it would smoothly available in the KB for any further enhanced searching or reasoning services, which would enables specific context awareness features.

Thus, if on one hand domain experts could at any time contribute to the update or insertion of new knowledge or semantics, on the other hand if there is any new component added to the system, or if any new standard or update happens, its related information could also be added to the meta-knowledge (ontology). Knowledge engineers accomplish these new updates or changes. Depending in the structure of the ontology and in how its related reasoning services are made, it would be possible that such changes requires updates on the services that uses the knowledge base to maintain its interoperable stage.

Both branches of this approach follow the MENTOR methodology as procedure to build the reference knowledge. The domain lexicon branch follows it till the thesaurus and the meta-knowledge till the ontology building steps. The process followed for the thesaurus building has, three steps: (1) terminology gathering; (2) glossary building; and (3) thesaurus building (Fig. 3).
Fig. 3

Thesaurus building methodology (Sarraipa et al. 2010)

The terminology-gathering step concerns the process of collecting all relevant terms or concepts in a specific domain previously defined. All the participants in the process should give their inputs. There is no rule from where the terms should come, since they are related to the domain established. All the concepts provided from contributors are acceptable in this step, as nobody has authority at the moment to erase another’s participant term. Thus, the terms should be collected with reference to the contributor, in order to enable each contributor to provide term’s definitions during the next step.

The Glossary building step builds a glossary in the domain defined. It starts with annotations attribution to the terms collected in the step before (Terminology gathering). Then, each contributor should provide the annotations for his/her own terms. After having all the terms provided with annotations, it proceeds to the terms revision cycle to reach a reference definition. The process for revision of terms can have four mismatches cases:
  • Existence of two syntactically different terms with the same meaning description—the action is to adopt one of the terms for being the reference in such semantics meaning.

  • Existence of two syntactical equal terms with the same meaning description—the action is to erase one of them.

  • Existence of syntactically different terms with two different meaning descriptions—no action needed, both must be kept.

  • Existence of two syntactically equal terms with two different meaning descriptions—the action is to consolidate all the provided descriptions together in one of them and erase the other. In such a case, a new term could be proposed to the list if there is no agreement in the conjunction of the input descriptions and if the term to be born is not present in the terminology list.

After a careful revision of all the terms with a successful agreement in their meaning consolidation, the glossary is defined from the terminology list in the domain specified.

The thesaurus-building step is composed of a cycle where the knowledge engineers and the domain experts define a taxonomic structure from the glossary terms, establishing some as thesaurus node terms (e.g. “Cantidad” or “Forma” in thesaurus presented at Fig. 5). Afterwards, the other terms are classified into semantic proper paths in the existent taxonomic structure, until reaching the thesaurus leafs. If there is an agreement in the structure and in the terms classified, the thesaurus is defined. If not, the cycle is repeated.

ALTER-NATIVA’s knowledge base

ALTER-NATIVA is an ALFA III project, which main goal is to promote higher Education in Latin America as a means to contribute to the economic and social development of the region (European Commision 2012). The main goals of the project were to: (1) provide education for everyone; (2) provide an environment of formation to professors when leading with persons with some disabilities; and (3) balance the inequalities of opportunities when accessing information.

In ALTER-NATIVA it was implemented a platform to support the project goals described above and in the establishment of an international network of higher education institutions with recognized expertise in the areas of pedagogical education and development of information technology. Such platform is composed by four main elements: (1) ATutor, (2) COLABORA, (3) e-Learning Repository, and (4) a reference KB.

Accessible Tutor well known as ATutor (2013) was the learning management system selected in the ALTERNATIVA project to support two main and necessary process for achieving the projects objectives: (1) be the learning environment for delivering to teachers different courses in order to improve their abilities to attend the diversity in the learning process; (2) be the environment for creating learning experiences which include the creation of accessible open educational resources.

The creation of accessible Open Educational Resources (OER) in ATutor is also supported by the integration of TinyMCE (2013). TinyMCE is the selected web content authoring tool, which provides a friendly user interface developed completely in JavaScript language as well as a set of functionalities that allow teachers to create web pages in an intuitive way without worrying about HTML code, because the editor automatically creates HTML. TinyMCE was improved in order to provide a better support when the teachers create accessible and OER; in particular, some accessibility issues were addressed in order to facilitate to teachers the attention of the Web accessibility Guidelines 2.0 in the OER creation process.

COLABORA is an infrastructure that facilitates the collaboration activities of the ALTER-NATIVA network. A community of practice is a group of people who share a concern or a set of problems of common interest about a topic and who deepen on their knowledge through on-going interaction. Thus, the COLABORA platform has allowed the management of communities of practice in favour of attention to diversity. In COLABORA, tasks are elements that allow the achievement of activities, which are part of the interaction between collaborators of each community of practice.

ALTER-NATIVA has an e-Learning Repository of accessible learning objects, which aims to organize, store and retrieve educational resources produced by the members of the ALTER-NATIVA network. The objects in the repository are organized into areas such as science, mathematics, language and communication. The repository has specific tools for labelling objects. However, it also allows import labelled object from the ATutor component. It uses the standard LOM and IMS AccessForAll (W3C 2008) to add accessibility information in learning objects. LOM is a multi-part standard that specifies Learning Object Metadata. The purpose of this multi-part standard is to facilitate search, evaluation, acquisition, and use of learning objects, for instance by learners or instructors or automated software processes (IEEE 2002).

The IMS AccessForAll intends to facilitate the formalization of Digital Resource Description (DRD) and Personal Needs and Preferences (PNP), through a meta-data specification. DRD is a lightweight metadata schema for describing and linking to digital objects. It is based on Qualified Dublin Core with local extensions. It is intended for use with simple digital objects as an alternative to more complex schemas (Natlib 2004). PNP is a meta-data that intends to specify how resources are to be presented and structured; how resources are to be controlled and operated; and, what supplementary or alternative resources are to be supplied. All of this has the main goal to meet the needs (preferences) of learners with disabilities and of anyone in a disabling context, with the purpose of offering them an appropriate interaction with digital resources, including configuration of assistive technologies.

In a kind of conclusion both DRD and PNP are parts of the ISO/IEC 24751 standard, which derives from IMS Learner Information Package Accessibility and IMS AccessForAll that intends to facilitate the matching of individual user needs and preferences with educational digital resources that meet those needs and preferences (Baldiris et al. 2011).

The establishment of the ALTER-NATIVA KB arises from the need to represent all the knowledge related with the project outcomes. Thus, it has three main objectives, the first is to enable professors to establish cooperation’s through COLABORA tool; the second is related to possibility to create Virtual Learning Objects (VLOs) in ATutor, which then would be saved and categorized using a common vocabulary (domain lexicon) understood by all the community; and finally, the third is related to the searching and recommendations of people (e.g. professors) or VLOs available at the e-Learning Repository related to specific domain topics characteristics. To accomplish this, the KB is provided with specific web services and graphic user interfaces to facilitate such mentioned features and a consistent knowledge management. The proposed knowledge management approach main purpose is to facilitate community members to actively participate in the constant domain knowledge updating process. Two distinct parts compose the ALTER-NATIVA’s KB: the “Domain Lexicon” and the “Meta Knowledge” (Fig. 4).
Fig. 4

An excerpt of the ALTER-NATIVA KB

ALTER-NATIVA’s domain lexicon

The “Domain Lexicon” part of the proposed KB is dedicated to represent the lexicon of ALTER-NATIVA. Thus, its main purpose is to have a set of reference concepts and meanings about its domain represented in thesauri. The ALTER-NATIVA thesauri represent the information related to the concepts of four distinct areas, namely Languages, Mathematics, Science, Teaching and Supporting Tools. The definition of each concept (keyword), the creation or agreement date and the author(s) of such information compose them. Each of the thesauri is available in a wiki kind of interface enable a public access to its contents. Figure 5 illustrates the main page of the Mathematics thesaurus, which shows its hierarchically representation and the access point to the other public available thesaurus.
Fig. 5

Mathematics thesaurus—hierarchical representation of its concepts

ALTER-NATIVA’s meta-knowledge

Six main classes, compose the “Meta Knowledge” branch of the ALTER-NATIVA KB, where the Users and VLOs classes are its core, because directly or indirectly they are related to all the other classes (Fig. 4). And mainly because, all the information and recommendations provided by this system have as target the elements characterized by both of these classes (people and VLOs).

The Users class represents all the stakeholders (people) of the ALTER-NATIVA community, namely their different characteristics, which are related to: (1) General; (2) Accessibility; and (3) ALTER-NATIVA profiles respectively. The General Profile class contains the general information related to a user, in particular to the communities, which a user belongs and also their level of studies accomplished. The Accessibility Profile class contains the information related to the disabilities of a user (e.g. visual; hearing; physical; cognitive). By the last, the Alternativa Profile class encompasses the information about the teaching profiles, namely the specialization area, the investigation groups, which a user has been involved, the level of knowledge when leading with some kind of disability, etc.

The Alternativa technologies class is conceived to represent the technologies information of the platform. It aggregates the representation of the VLO metadata, which integrates parts of the LOM and DRD. It is from these metadata elements that it is used, through instantiation the concepts represented in the thesauri, which are used to characterize the VLOs. In the Recommendations section are represented the interactions between users, namely the activities that they share and the VLOs that they accessed or consulted. Finally, the Preferences section has the information related with users preferences about how they want to access to the information.

The Support Tools class is designed to represent the supporting tools used in the learning executions and its characteristics. One tool example is shown in Table 1. As can be observed, it contains the information related to the kind of users for which a support tool is designed to. This class is connected to the thesaurus through keywords of the supporting tool description.
Table 1

Support tools description

Finally, the performance class represents all the performance scenarios (e.g. learning implementations) using ALTER-NATIVA platform, namely through COLABORA, and the user’s participations at these performance scenarios.

Knowledge base management interfaces

The ALTER-NATIVA KB was developed with Protégé-OWL editor. It was decided to use the OWL 1.0 standard to represent ontologies. The ALTER-NATIVA KB is managed at two different levels accordingly to the knowledge management approach proposed. As a consequence there are two main user interfaces to edit: (1) the meta-knowledge; and (2) the thesauri (Fig. 6).
Fig. 6

KB management interfaces

The meta-knowledge is managed by knowledge engineers (ontology managers), which can directly edit the ontology through the protégé editor tool. In this case, it was set up two solutions, one offline and another online. The offline way is through the protégé regular tool that connects directly to the ALTER-NATIVA server, which then can edit and upload changes in the OWL file. The online version uses the WebProtégé plug-in, which is a Collaborative Ontology Editor and Knowledge Acquisition Tool for the Web. WebProtégé includes a set of predefined tabs, which contain the most popular functionality in the protégé desktop editor (Noy et al. 2000).

The thesauri are available through a customized wiki. Since a thesaurus is like a taxonomy of domain concepts composed by its reference meanings, it was used MediaWiki as a front-end to enable communication and knowledge gathering from the ALTER-NATIVA community domain experts. In addition, due to its capability of providing a dynamic view of the wiki’s category structure as a tree, it was used the CategoryTree extension on the Media Wiki. As a result it was possible to represent a hierarchical representation of the concepts used in taxonomies. Figure 5 illustrates the thesaurus of Mathematics, which is available at the URL: http://alternativathesaurus.udistrital.edu.co/alternativa/matematica/?title=Página_principal. Through each one of the wiki Thesaurus main page is possible to access the other ALTER-NATIVA public wiki thesaurus.

With such kind of wiki-based user interfaces it was possible to provide users (domain experts) the possibility of editing, updating and creating concepts, which facilitates the maintenance of this community knowledge alive. The thesaurus in the wiki was configured to have three kinds of users: (1) Bot; (2) Administrator; and (3) Bureaucratic. The Bot user is only able to consult information. The thesaurus Administrator is able to manage the information represented in the thesaurus. There are three main actions that the Administrator can do to keep the thesaurus updated: (1) create a new concept; (2) edit a concept; and (3) delete a concept. The Bureaucratic user can execute two different actions: (1) delete a user from the database; and (2) promote a user to Administrator.

As presented in the knowledge management approach, it is required to align the knowledge of the community (domain experts) with the knowledge represented in the system. Thus, if there is an update made by the domain experts in the thesauri, it should be smoothly updated in the KB in order to be right way included in the searching or reasoning services outputs. The interoperability establishment between the wiki Data Base (DB) and the ALTER-NATIVA KB require a synchronization interface. The proposed synchronization functionality uses the wiki DB to detect any changes that have occurred since the last run (verification), to then update the KB accordingly. In the following subsections are explained in detail three actions that represent thesaurus typical changes: (1) concept deletion; (2) concept creation; and (3) concept edition.

Delete a concept

The elimination of a concept directly affects the structure of the thesaurus tree. If the deleted concept does not have children associated, it is just removed from the thesaurus tree at the ALTER-NATIVA KB. But if the deleted concept have children associated it is necessary to take some actions to reflect the wiki status in the KB. In this case, the children of the deleted concept are positioned in the thesaurus root. In Fig. 7, it is shown the initial and final status of the thesaurus tree after the deletion of a concept with children. As can be seen, the example illustrated is related to the deletion of the “Concept b”, which in the initial thesaurus tree had the children “Concept c” and “Concept d”. In the final state is possible to see that these concepts are shifted to the root (Thing).
Fig. 7

Delete of a concept with children associated

Create a concept

As in the concept deletion action, the creation of a concept directly affects the structure of the thesaurus tree. So it is necessary to rearrange the three to handle the three cases that are explained at the following.

The Fig. 8 illustrates the case where the inserted concept (“Concept h”) has another existing concept (“Concept b”) defined as its master (parent) concept. In this case, the action to be taken is to insert the “Concept h” has child of “Concept b”.
Fig. 8

Creation of a concept that points to an existing master concept

The Fig. 9 illustrates the case where the inserted concept (“Concept h”) has (by mistake) a non-existing concept (“Concept z”) defined as its master (parent) concept. In this case, since the “Concept z” does not exist in the initial state of the thesaurus tree, the action to take is to insert the “Concept h” has child of the “Thing” class (to put it in the root).
Fig. 9

Creation of a concept that points to a non-existing master concept

It is also possible to occur the case where the inserted concept (“Concept y”) was already defined as a master concept of an already concept existent in the root (“Concept f”). Fig. 10 illustrates this mentioned case. When this new concept (“Concept y”) is inserted in the tree, “Concept f” will be placed bellow it.
Fig. 10

The creates concept is referred as a master concept

Figure 10 also illustrates a case where two atomic cases occurred simultaneously. Table 2 contains a summary explanation of all the possible conjugations of the mentioned atomic situations and which are the steps to execute by the synchronization module.
Table 2

Use case of concept creation

Inserted master concept exists?

Is there a concept in the root, which has defined the new introduced concept as its master (parent)?

Action

Yes

Yes

Put new concept below the class referred as master concept

Put the class tree of the concept that is in the root and which refers the new concept as master concept, below the new introduced concept

No

Put new concept below the concept referred as master concept

No

Yes

Put new concept below the class thing

Put the class tree of the concept that is in the root and which refers the new concept as master concept, below the new introduced concept

No

Put new concept below the class “Thing” (root)

Edit a concept

When a concept is edited, if the master concept field of the form is not changed, it is only necessary to save (update) the new information in the KB Thesaurus. If the master concept is changed, may occur two situations: (1) the new master concept does not exist in the Thesaurus tree, so is necessary to execute the steps illustrated on Fig. 9; or (2) the new master concept exists in the tree and is necessary to execute the steps illustrated on Fig. 8.

E-learning knowledge-based services

The authors presented a specific knowledge management approach, which facilitates a formalised knowledge organisation able to afford learners or enterprises trainees, with efficient knowledge transfer instruments supported by advanced services (e.g. e-Learning objects recommendations; adaptable e-training programmes).

ALTER-NATIVA knowledge base services

The ALTER-NATIVA KB web services are responsible for giving worldwide users the mechanism of Knowledge Management and reasoning trough the Web. These services use JENA libraries, a java API for managing OWL ontologies. The developed services are classified in two types: the KB and Recommendation services.

KB services are of two groups: (1) Knowledge Management Services, the ones related to the knowledge edition; (2) Knowledge Query Services are the ones related to the knowledge reasoning. All these services are related to the VLO Consulted, User Model (profiles), Performance Scenarios, Thesaurus, Users Interactions and in general way with VLO (Fig. 11).
Fig. 11

Knowledge management services

The User Interactions and the VLO Consulted are a kind of a log of the system, which recorded information is later used to perform recommendations. Figure 11 also presents how the KB services are integrated with the other ALTER-NATIVA platform elements. The blue colour represents the connection between the ATutor and the ALTER-NATIVA KB. It illustrates that ATutor interacts with VLOs, illustrating its creation and edition, which uses the keywords available in the Thesaurus for its characterization. The orange colour represents the connection between the e-Learning Repository (“Repositorio”) and the necessary services able to upload the metadata information of the VLOs into the KB. The red colour represents the services that COLABORA can use to manage the information related to the users and their profiles. Is also represents the interactions between the users and their participations in “Performance Scenarios”. The green colour represents the synchronization feature, which was developed in a form of web service, between the Web Thesaurus (wiki) and its representation in the ALTER-NATIVA KB (Thesaurus).

Recommendation services

The recommendation services have as goal to support the searching of VLOs and users. The VLOs could be related to different characteristics as for instance, about specific trainees diversities. The users searching feature be related to the searching of professors with a specific profile (e.g. teaching experiences).

The recommendations are essentially done based on the User Model (profile), VLOs metadata, log information of the consulted VLO and users’ interactions. This information is introduced by the system in the KB enabling the analysis of various patterns for specific items recommendations purposes. In ALTER-NATIVA the output of the developed recommendation services is for ATutor and COLABORA consume. ATutor uses recommendation services to suggest to users potential useful VLOs in the development of courses, or other VLOs. COLABORA mainly uses these services to suggest people to interact with in a specific activity.

Recommendation of users concerns to the automatic suggestion of users to interact with, based on: (1) similar teaching experience; (2) a specific similar profile characteristics; (3) number of interactions between two determined users; (4) number of common profile characteristics between two users; and also (5) a combination of number of interactions with common user profiles.

When the recommendations deal with more than one characteristic, the process of choosing the most appropriated item is by calculating the one that has more similarities to the set of chosen target characteristics. Thus it was developed a function to be used in such process, which calculates the strength of each item with potential to be recommended.

The services of recommendation of VLOs are based on: (1) higher rating; (2) most used; (3) number of similarities with other VLOs consulted by the user; and (4) combination of the previous patterns.

To automatically suggest a VLO is necessary to have a log record of the consulted VLOs. Such information is formally represented in the KB as a table (Table 3), which interrelates the VLOs with the user identifiers, the number of times each user consulted the VLO and his/her associated evaluation rating about such VLO (1—Good or 0—Bad).
Table 3

Log information associated to a VLO

VLO identifier

User identifier

Number of times consulted

Rating

In the top left of Fig. 12, it is shown how the algorithm for suggestion of similar VLOs based on its characterization keywords. It starts by collecting the identifiers of all the VLO consulted by a user. Then, are collected the keywords that characterize those VLOs (top right). After this, as can be observed in the middle left of the figure, it is generated a new table without the VLO consulted by the user. It is used to collect the VLO with equal keywords to be then counted the number of keywords in common. Finally, the recommendation result is represented by the bottom table, which presents the VLO with common keywords, ordered by the high number of keywords in common.
Fig. 12

Algorithm to obtain the most similar VLOs accordingly to the user “1” logs

The combined recommendation of VLOs is done based on all the previous described recommendation services for VLO. Thus, this service returns the VLO with the higher strength based on the conjunction of the various evaluation patterns characteristics as number of times consulted, rating and number of similarities.

An advanced e-training service for industrial trainees

The presented ALTER-NATIVA knowledge management approach was already experimented in industrial scenarios, demonstrating its full appropriateness and efficiency in supporting other environments’ needs. The followed MENTOR knowledge building process to build a reference lexicon in ALTER-NATIVA was already tested in furniture and mechanical industrial environments respectively, as presented in Jardim-Goncalves et al. (2011) and Sarraipa et al. (2010). On the other hand, the used ontology-based solution to represent the ALTER-NATIVA e-Learning knowledge, which supports the described services implementation, was also used in a similar way, by the authors, for industrial purposes. It was used to implement an adaptable training programme service for automotive and construction industries, which training materials resulted from the CoSpaces research project. CoSpaces was a project funded by the EC under the IST Programme of the FP6, which overall objective was to develop organisational models and distributed technologies supporting innovative collaborative workspaces for individuals and project teams within distributed virtual manufacturing enterprises (Maló et al. 2008; Sarraipa et al. 2012). Training in CoSpaces project aimed exactly at providing knowledge and skills that allow key personnel within distributed manufacturing enterprises to understand collaborative practices and acquire the practical experience of collaborative design engineering methods, supported on meaningful case studies and demonstrators (Sarraipa et al. 2009, 2012).

Consequently, an ontology to represent the CoSpaces training to facilitate the categorization of its elements and subsequently the reasoning over them was developed. It focused in he Training in Collaborative Working (TiCW) curriculum (Fig. 13), which follows an integrative structure (matrix) established over two-dimensional axes: horizontally composed by training levels; and vertically by reference content areas. Each of such levels and content areas has specific courses (listed in the right part of Fig. 13), and each course has a set of training modules (bottom part of Fig. 13).
Fig. 13

TiCW training curriculum and its composition elements overview

All the TiCW training courses have a synopsis defined to represent its related information like: training objectives; precedencies, keywords, and other metadata. Figure 14 presents an excerpt of this use case e-training ontology. It illustrates specifically the knowledge related to one of the modelled courses—the CUI (CoSpaces User Interfaces) course. It also shows that the CUI course has 3 modules, and it is linked to the Collaborative Technologies curriculum topic area. The recommended precedence’s and the level of training to which the CUI course belongs are also in the model (Fig. 14), among other relevant characteristics.
Fig. 14

Training system knowledge base

Due to the existence of the training curriculum matrix and related data representation in an ontology, it is possible to reason over it to generate specific training programmes, which contents (courses / modules) are presented accordingly (adapted) to the user needs and following the synopsis pre-determined pedagogical directives (e.g. precedencies).

The developed adaptable training programme service is available at http://gris-public.uninova.pt:8080/cospaces/ATPS_in.jsp (website Accessed at March 2014), in a form of a web page. It presents a set of keywords able to be selected to generate on the fly adapted training programmes related to the choices.

The training impact of these specific programmes is proportionally related to the ability, skills and competences that such training could give to an worker, which could directly influence on a job performance, such as operations, human resources policies, or management and leadership (Nielson 2009). Online training (e-training) is a major driver to promote the development of competencies and knowledge in enterprises. A lack of adaptable e-training programmes based on trainees’ profiles or needs and a continuous maintenance of the training materials prevents the sustainability of industrial training deployment.

Conclusions

In this research report it is presented a KB developed to represent the ALTER-NATIVA’s knowledge. It followed a taxis path to build knowledge and fits as the central element in the proposed Knowledge Management approach. Maedche & Staab stated that by defining shared and common domain theories, ontologies help both people and machines to communicate concisely, supporting the exchange of semantics and not only syntax (Maedche and Staab 2001). In align with this, the ontology, to which the developed KB relates to, provide a lexicon and its meanings that describe the ALTER-NATIVA system and domain. It formally represents not only the domain knowledge, but also the system characteristics and objectives in the support of the ALTER-NATIVA’s network establishment. Therefore, it integrates technical solutions able to harmonize a community knowledge view providing a semantic interoperable basis to smoothly support the creation of virtual collaborative communities and resources, which has been practiced in the ALTER-NATIVA community.

It was also noticed that this kind of knowledge organization systems enables context awareness abilities. This has been reached through available recommendation services. The knowing how users (i.e. professors and trainees) and their profiles are related with the VLOs and what kind of tools that can be used to reduce the learning vulnerability (diversity) or to enhance accessibility of trainees, is a contextual information that the system could be aware to react accordingly. In this case, a better suggestion on the training implementation (e.g. an appropriate VLO suggestion) could be done by the system.

Additionally, an e-Training application case was described to demonstrate the applicability of the presented solutions or approaches in industry specific purposes, which demands for effective knowledge transfer to enhance the delivery of skills and competences in the workers. Thus, despite to a direct contribution with a semantic interoperability resolution (in the common lexicon establishment), the proposed knowledge management approach, also has the potential to contribute to an efficient training implementation, which training programmes as in the presented industrial case study could be related to the establishment of interoperable and collaborative technological solutions.

Future work

Authors intend to integrate the proposed knowledge management approach in the furniture industry under the IMAGINE project (IMAGINE Consortium 2013). It would be done by develop a wiki for supporting furniture domain experts in their domain lexicon establishment. Such lexicon would be also further integrated in a KB to be used by enhanced business services developed on that project. These services would be able to contextualize, as an example, in appropriate suggestions of effective manufacturers for a specific product.

Additionally, authors would like to introduce diversity characteristics in the training approaches in their future training implementations. This intends to be aligned with the statement: “A sustainable culture that promotes a respectful, inclusive, knowledge-based environment within which each person has the chance to develop their potential and contribute to the success of the organization” (Impact International 2013).

Notes

Acknowledgments

The research leading to these results has received funding from the European Union under grant agreement: ALTER-NATIVA Project (DCIALA/19.09.01/10/21526/245-575/ALFA III (2010) 88). Authors would like to thank all the participants in the project that would have contributed to this work results. Additionally, they are highly indebted to CALE, CAM and CA Communities for their contributions in the domain lexicon building. Finally, authors would also like to thank to all involved in the activities supporting the development of the CoSpaces Training System, namely the CoSpaces project community/members.

References

  1. ATutor. (2013). ATutor: Learning management system. Retrieved from the web at November 2013: http://atutor.ca/atutor.
  2. Baldiris, S., Avila, C., Rivera, P. A., Guevara, J. C., & Fabregat, R. (2011). Web editing module for tagging metadata of the Fedora Commons repository learning objects under DRD and LOM standards. In 2012 Frontiers in education conference proceedings (Vol. 0, pp. S2E–1–S2E–5). doi:10.1109/FIE.2011.6142750.
  3. Bhattacharya, A., Tiwari, M. K., & Harding, J. A. (2010). A framework for ontology based decision support system for e-learning modules, business modeling and manufacturing systems. Journal of Intelligent Manufacturing, 23(5), 1763–1781. doi:10.1007/s10845-010-0480-6.CrossRefGoogle Scholar
  4. Camarinha-Matos, L. M., & Afsarmanesh, H. (2007). A comprehensive modeling framework for collaborative networked organizations. Journal of Intelligent Manufacturing, 18(5), 529–542. doi:10.1007/s10845-007-0063-3.CrossRefGoogle Scholar
  5. Chawhan, A. K. (2012). Knowledge management and e-learning. In eCommunications—CSI Communications-Knowledge Digest for IT Community; ISSN 0970–0647X; Vol. 36; Issue 3; June 2012.Google Scholar
  6. Chu, M. T., Khosla, R., & Nishida, T. (2010). Communities of practice model driven knowledge management in multinational knowledge based enterprises. Journal of Intelligent Manufacturing, 23(5), 1707–1720. doi:10.1007/s10845-010-0472-6.CrossRefGoogle Scholar
  7. Costa, R., Lima, C., Sarraipa, J., & Jardim-Gonçalves, R. (2013). Facilitating knowledge sharing and reuse in building and construction domain: An ontology-based approach. Journal of Intelligent Manufacturing. doi:10.1007/s10845-013-0856-5.
  8. European Commision. (2012). Development and cooperation—EUROPEAID. Retrieved at October 17, 2013, from http://ec.europa.eu/europeaid/where/latin-america/regional-cooperation/alfa/publications_en.htm.
  9. Guarino, N., & Giaretta, P. (1995). Ontologies and knowledge bases: Towards a terminological clarification. Towards Very Large Knowledge Bases: Knowledge Building and Knowledge Sharing, 25–32. Retrieved from the web at October 2013: http://www.csee.umbc.edu/771/papers/KBKS95.pdf.Z.
  10. IEEE. (2002). Draft standard for learning object metadata. Retrieved from web at October 2013: http://ltsc.ieee.org/wg12/files/LOM_1484_12_1_v1_Final_Draft.pdf.
  11. IMAGINE Consortium. (2013). IMAGINE—Innovative end-to-end Management of Dynamic Manufacturing Networks. Available in the web at December 2013: http://www.imagine-futurefactory.eu.
  12. Impact International. (2013). Diversity & Inclusion. Retrieved from the web at December 2013: http://www.impactinternational.com/diversity-inclusion.
  13. Jardim-Goncalves, R., Sarraipa, J., Agostinho, C., & Panetto, H. (2011). Knowledge framework for intelligent manufacturing systems. Journal of Intelligent Manufacturing, 22(5), 725–735. doi:10.1007/s10845-009-0332-4.CrossRefGoogle Scholar
  14. Karayel, D., Sancak, S., & Keles, R. (2004). General framework for distributed knowledge management in mechatronic systems. Journal of Intelligent Manufacturing, 15(4), 511–515. Retrieved from http://dblp.uni-trier.de/db/journals/jim/jim15.html.
  15. Maedche, A., & Staab, S. (2001). Ontology learning for the semantic web. IEEE Intelligent Systems, 16(2), 72–79. doi:10.1109/5254.920602.CrossRefGoogle Scholar
  16. Maló, P., Sarraipa, J., Jardim-Gonçalves, R., & Steiger-Garção, A. (2008). The CoSpaces training system. In ICE2008 conference proceedings—14th international conference on concurrent enterprise, 23–25 Jun 2008, Lisbon, Portugal.Google Scholar
  17. May, M. (2002). Business process management: Strategic integration in a web-enabled environment (Management Briefings Executive Series). Financial Times Management. Retrieved from http://www.amazon.com/exec/obidos/redirect?tag=citeulike07-20&path=ASIN/0273661086.
  18. Natlib. (2004). Digital resource description (DRD) application profile. Retrieved from the web at October 2013: http://archive.today/fdzIa.
  19. Nielson, B. (2009). Relevance of learning versus relevance of training and development; http://ezinearticles.com/?Relevance-of-Learning-Versus-Relevance-of-Training-and-Development&id=2255045.
  20. Noy, N. F., Sintek, M., Decker, S., Crubézy, M., Fergerson, R. W., & Musen, M. A. (2000). Creating semantic web contents with protégé-2000. IEEE Intelligent Systems, 16(2), 60–71. doi:10.1109/5254.920601.
  21. O’Leary, D. E. (1998). Using AI in knowledge management: Knowledge bases and ontologies. Intelligent Systems and their Applications, IEEE, 13(3), 34–39. doi:10.1109/5254.683180.
  22. Rhoads, E. (2002). Knowledge management and e-learning are two sides of the same coin. Retrieved from the web at December 2013: http://www.academia.edu/3656488/KM_e_Learning_Elsa_Rhoads.
  23. Rolstadås, A. (2011). Experience from continuing education using e-learning. Journal of Intelligent Manufacturing, 24(3), 511–516. doi:10.1007/s10845-011-0542-4.CrossRefGoogle Scholar
  24. Roy, A., & Raymond, L. (2008). Meeting the training needs of SMEs: Is e-learning a solution? The Electronic Journal of e-Learning, 6(2), 89–98.Google Scholar
  25. Ruiz, J. G., Mintzer, M. J., & Leipzig, R. M. (2006). The impact of e-learning in medical education. Academic Medicine: Journal of the Association of American Medical Colleges, 81(3), 207–212. Retrieved from http://view.ncbi.nlm.nih.gov/pubmed/16501260.
  26. Sarraipa, J. (2012). PhD Thesis: Semantic adaptability for the systems interoperability. Faculdade de Ciências e Tecnologia, New University of Lisbon.Google Scholar
  27. Sarraipa, J., Baldiris, S., Fabregat, R., & Jardim-Goncalves, R. (2012). Knowledge representation in support of adaptable eLearning services for all. In Proceedings of the 4th international conference on software development for enhancing accessibility and fighting info-exclusion (DSAI 2012).Google Scholar
  28. Sarraipa, J., Jardim-Gonçalves, R., & Steiger-Garção, A. (2010). MENTOR: An enabler for interoperable intelligent systems. International Journal of General Systems, 39(5), 557–573. Retrieved from http://dblp.uni-trier.de/db/journals/ijgs/ijgs39.html#SarraipaJS10.
  29. Sarraipa, J., Malo, P., Patel, H., Pettitt, M., Hardiman, S., Banassino, M., Syllignakis, A., & Carter, M. (2009). D102—training strategy and plan. A CoSpaces project deliverable. 30th April 2009.Google Scholar
  30. Sarraipa, J., Gomes-de-Oliveira, P., Marques-Lucena, C., Jardim-Gonçalves R., & Silva, J. M. (2013). E-training development approach for enterprise knowledge evolution. In The ASME international mechanical engineering congress and exposition, IMECE 2013, held at November 15–21, 2013, San Diego, CA, USA.Google Scholar
  31. thefreedictionary.com. (2003). Taxis definition. Retrieved from the web at October 2013: http://www.thefreedictionary.com/taxis.
  32. TinyMCE. (2013). Accessibility with TinyMCE. Retrieved from the web at October 2013: http://www.tinymce.com/wiki.php/TinyMCE3x:Accessibility.
  33. UCSF. (2012). Chapter 12: Managing Diversity in the Workplace. Retrieved from the web at October 2013:  http://ucsfhr.ucsf.edu/index.php/pubs/hrguidearticle/chapter-12-managing-diversity-in-the-workplace/.
  34. University of Michigan EECS Department. (2013). A survey of cognitive and agent architectures. Retrieved from the web at December 2013: http://ai.eecs.umich.edu/cogarch0/common/prop/metaknow.html.
  35. Wikipedia.org. (2013). Taxonomy definition. Retrieved from the web at October 2013: http://en.wikipedia.org/wiki/Taxonomy.
  36. W3C. (2008). Web Content Accessibility Guidelines (WCAG) 2.0. Retrieved from web at December 2013: http://www.w3.org/TR/WCAG20/.

Copyright information

© Springer Science+Business Media New York 2014

Authors and Affiliations

  • Joao Sarraipa
    • 1
    • 2
  • Catarina Marques-Lucena
    • 1
    • 2
  • Silvia Baldiris
    • 3
  • Ramón Fabregat
    • 3
  • Silvana Aciar
    • 4
  1. 1.DEE/FCTUniversidade Nova de LisboaCaparicaPortugal
  2. 2.UNINOVA-GRIS, Centre of Technology and SystemsCaparicaPortugal
  3. 3.Institute of Informatics and Applications (IIiA)University of GironaGironaSpain
  4. 4.Universidade Nacional de San JuanSan JuanArgentina

Personalised recommendations