Mike Salampasis - Academia.edu (original) (raw)

Papers by Mike Salampasis

This paper discusses the current status about the adoption of Information Technology (IT) by the ... more This paper discusses the current status about the adoption of Information Technology (IT) by the extension services in a developing country such as Albania. Similar to other former Easter European countries, Albania is moving towards a western-type economy and an intensive effort is made to promote the use of IT for providing decentralised extension services of high quality. This is considered as a key factor in order to increase efficiency and effectiveness of the agricultural industry. In this paper we also report preliminary results of an ongoing research and development programme which is financially supported from the European Union (EU). The main objective of this programme is to set up information centers for "one stop" information shopping in different areas of Albania. Since this programme involves the largest public agricultural organisations (ministry of agriculture and food), the largest private associations in Albania as well as academia, we believe that exper...

Lecture Notes in Computer Science, 2005

This paper reports the results of a user-centered experiment which examined the effect of paralle... more This paper reports the results of a user-centered experiment which examined the effect of parallel multi-database searching using automated collection fusion strategies on information seeking performance. Three conditions were tested in the experiment. Subjects in the first condition performed search tasks in a WWW-based distributed hypermedia digital library which did not support parallel, concurrent searching of multiple collections, and did not offer any automated mechanism for source selection. Subjects in the second and the third conditions performed parallel multi-database search tasks in the same library with the support of two automated collection fusion strategies (uniform and link-based), each solving the collection fusion problem using a different approach. The results show that information-seeking performance tends to be positively affected when the eclectic link-based method was used. On the other hand, the uniform collection fusion method which treats all the sub-collections in the same manner, does not present any benefit in comparison to information seeking environments in which users must manually select sources and parallel multi-database searching is not provided.

Lecture Notes in Computer Science, 2012

Pseudo-relevance feedback (PRF) is an effective approach in Information Retrieval but unfortunate... more Pseudo-relevance feedback (PRF) is an effective approach in Information Retrieval but unfortunately many experiments have shown that PRF is ineffective in patent retrieval. This is because the quality of initial results in the patent retrieval is poor and therefore estimating a relevance model via PRF often hurts the retrieval performance due to off-topic terms. We propose a learning to rank framework for estimating the effectiveness of a patent document in terms of its performance in PRF. Specifically, the knowledge of effective feedback documents on past queries is used to estimate effective feedback documents for new queries. This is achieved by introducing features correlated with feedback document effectiveness. We use patent-specific contents to define such features. We then apply regression to predict document effectiveness given the proposed features. We evaluated the effectiveness of the proposed method on the patent prior art search collection CLEF-IP 2010. Our experimental results show significantly improved retrieval accuracy over a PRF baseline which expands the query using all top-ranked documents.

2008 Panhellenic Conference on Informatics, 2008

Intelligent adaptive user interfaces have been long explored as a promising technique to solve pr... more Intelligent adaptive user interfaces have been long explored as a promising technique to solve problems due to the complexity of modern user interfaces. These problems can occur in various application settings such as accessing the Web from small screen devices, over a telephone based interface or disabled people accessing the WWW using specialised voice web browsers. In this paper we present adaptive browsing shortcuts (ABS), a user interface personalisation mechanism that can be applied in the WWW and which is based on simple and continuous user modelling and monitoring. We discuss various personalisation methods that can be developed upon the idea of ABS and we present an application which utilises browsing shortcuts to increase navigability and information seeking performance of blind people using voice Web browsers. Results from user studies are reported which consistently indicate that ABS could provide effective user interface personalization based on solid user and interaction modelling.

Proceedings of the 2nd ACM workshop on Improving non english web searching, 2008

Greek is one of the most difficult languages to handle in Web Information Retrieval (IR) related ... more Greek is one of the most difficult languages to handle in Web Information Retrieval (IR) related tasks. Its difficulty stems from the fact that it is grammatically, morphologically and orthographically more complex than the lingua franca of IR, English. In this paper, we address a significant number of issues that originate from the Greek language. We use a number of techniques to determine the correct encoding that is used by web pages written in Greek. We test the effect of using a Greek stopword list in a realistic and controlled Web environment. We employ a character mapping scheme, in order to overcome the problem of the diversity of diacritics used in the language, such as accents and diaeresis. We utilize word distance and fuzzy similarity metrics in order to make up for the different forms that nouns, verbs and articles appear because of conjugations and inflections and additionally handle greeklish queries, a transliterated form of Greek. The conducted experiments present some effective ways to increase the accuracy in Greek IR tasks.

Lecture Notes in Computer Science, 2009

Source selection deals with the problem of selecting the most appropriate information sources fro... more Source selection deals with the problem of selecting the most appropriate information sources from the set of, usually non-intersecting, available document collections. On the other hand, data fusion techniques (also known as metasearch techniques) deal with the problem of aggregating the results from multiple, usually completely or partly intersecting, document sources in order to provide a wider coverage and a

This paper discusses the use of linear programming in a decision-making system for broiler enterp... more This paper discusses the use of linear programming in a decision-making system for broiler enterprises. The general model of linear programming is presented and the specific conditions and constraints of a modern typical broiler enterprise are examined. A scenario of a case study that was applied in a collaborating broiler enterprise is discussed and the benefits of using linear programming

Proceedings of the sixteenth ACM conference on Hypertext and hypermedia, 2005

The WWW is today the biggest source of information and an essential tool for many activities of d... more The WWW is today the biggest source of information and an essential tool for many activities of daily life. Unfortunately, information seeking in this complex hypermedia environment is generally not an easy task. The potentially complex task of information seeking in the WWW is further complicated when the end-user is blind or visually impaired (VI). Usually, web pages are created

Universal Access in the Information Society, 2007

... of blind people in the WWW Christos Kouroupetroglou Æ Michail Salampasis Æ Athanasios Manitsa... more ... of blind people in the WWW Christos Kouroupetroglou Æ Michail Salampasis Æ Athanasios Manitsaris ... M. Salampasis Department of Informatics, ATEI of Thessaloniki, Thessaloniki 57 400, Greece e-mail: cs1msa@it.teithe.gr ...

New Review of Hypermedia and Multimedia, 2008

The World Wide Web is today the largest information seeking environment. Millions of people use i... more The World Wide Web is today the largest information seeking environment. Millions of people use it to satisfy their information needs. Although it is quite easy for able-bodied users to use it, there are still a lot of problems for people with disabilities. A major group of them are blind users. Blind users navigate the web in a different and

Information Processing & Management, 2008

ABSTRACT The problem of results merging in distributed information retrieval environments has gai... more ABSTRACT The problem of results merging in distributed information retrieval environments has gained significant attention the last years. Two generic approaches have been introduced in research. The first approach aims at estimating the relevance of the documents returned from the remote collections through ad hoc methodologies (such as weighted score merging, regression etc.) while the other is based on downloading all the documents locally, completely or partially, in order to calculate their relevance. Both approaches have advantages and disadvantages. Download methodologies are more effective but they pose a significant overhead on the process in terms of time and bandwidth. Approaches that rely solely on estimation on the other hand, usually depend on document relevance scores being reported by the remote collections in order to achieve maximum performance. In addition to that, regression algorithms, which have proved to be more effective than weighted scores merging algorithms, need a significant number of overlap documents in order to function effectively, practically requiring multiple interactions with the remote collections. The new algorithm that is introduced is based on adaptively downloading a limited, selected number of documents from the remote collections and estimating the relevance of the rest through regression methodologies. Thus it reconciles the above two approaches, combining their strengths, while minimizing their drawbacks, achieving the limited time and bandwidth overhead of the estimation approaches and the increased effectiveness of the download. The proposed algorithm is tested in a variety of settings and its performance is found to be significantly better than the former, while approximating that of the latter.

Thesis (Ph. D.)--University of Sunderland, 1997.

24th Pan-Hellenic Conference on Informatics

We present a graph-based approach for the data management tasks and the efficient operation of a ... more We present a graph-based approach for the data management tasks and the efficient operation of a system for sessionbased next-item recommendations. The proposed method can collect data continuously and incrementally from an ecommerce web site, thus seemingly prepare the necessary data infrastructure for the recommendation algorithm to operate without any excessive training phase. Our work aims at developing a recommender method that represents a balance between data processing and management efficiency requirements and the effectiveness of the recommendations produced. We use the Neo4j graph database to implement a prototype of such a system. Furthermore, we use an industry dataset corresponding to a typical e-commerce session-based scenario, and we report on experiments using our graph-based approach and other state-of-the-art machine learning and deep learning methods.

Searching for patents is usually a recall-oriented problem and depending on the patent search typ... more Searching for patents is usually a recall-oriented problem and depending on the patent search type, quite often a problem which is characterized by uncertainty and evolution or change of the information need. We propose an exploratory strategy for patent search that exploits the metadata already available in patents in addition to the results of clustering and entity mining that are performed at query time. The results (metadata, clusters and entities grouped in categories) can complement the ranked list of patents produced from the core search engine with useful information for the user (e.g. providing a concise overview of the search results) which are further exploited in a faceted and sessionbased interaction scheme that allows the users to focus their searches gradually and to change between search methods as their information need is better defined and their understanding of the topic evolves in response to found information. In addition, we propose the exploitation of Linked ...

One of the biggest issues the World Wide Web (WWW) community has to overcome nowadays is accessib... more One of the biggest issues the World Wide Web (WWW) community has to overcome nowadays is accessibility for all. The rapid development of the WWW using doubtful web authoring practices, together with the domination of the desktop metaphor in the web page design, created accessibility problems for people with disabilities using the WWW. Until now several solutions have been suggested to ease the accessibility problem. Some of them exploit the opportunities that the new idea of Semantic Web. In this paper an application framework is presented and discussed (together with key technologies of the Semantic Web), which is used to enhance the information seeking of blind users. The application framework makes use of the idea of the Semantic Web as a potential solution for helping blind users in their information seeking process in the WWW using metadata. Most important, the application framework could be generalized and used for different application domains in which metadata could be produ...

Proceedings of the 2008 ACM workshop on Large-Scale distributed systems for information retrieval, 2008

In this paper, a new source selection algorithm for uncooperative distributed information retriev... more In this paper, a new source selection algorithm for uncooperative distributed information retrieval environments is presented. The algorithm functions by modeling each information source as an integral, using the relevance score and the intra-collection position of its sampled documents in reference to a centralized sample index and selects the collections that cover the largest area in the rank-relevance space. Based

2007 IEEE 23rd International Conference on Data Engineering Workshop, 2007

Browsing shortcuts is a mechanism which facilitates blind people to move efficiently to various e... more Browsing shortcuts is a mechanism which facilitates blind people to move efficiently to various elements of a web page (e.g. functional elements such as forms, navigational aids etc.), hence operating effectively as an interaction method and a vital counterbalance to low accessibility of web pages. Results of a quantitative analysis which measured navigation performance and cognitive overhead criteria (task completion time, number of keystrokes, web page reading times) with and without the use of browsing shortcuts, showed that browsing shortcuts has a statistically significant positive effect on navigation performance of blind people. In this paper we examine further the idea of browsing shortcuts by presenting a personalised user interface of a specialised voice web browser. Three ways of personalising the user interface are presented based on the reordering and adaptation of browsing shortcuts and well as by incorporating recommendations about browsing shortcut selection.