“As far as public accessibility is concerned, one has to wonder what the point is of putting more stuff online when we do such a bad job of helping people to explore what is already there” (Norris, 2010). The quote sums up one of the problems that arises when dealing with the information search within the vastness of the web. The best known of research methods is undoubtedly the one “by keyword”. This method has limitations such as: the difficulty in the formulation of the "query" research, the problems concerning the use of different but overlapping vocabularies, the inability to formulate requests terminologically strong for knowledge domains not known. To overcome these obstacles have created new and more efficient systems of classification and search of information. Among these systems faceted search (from "facets": "aspects") is increasingly asserting and gives the possibility to create a data set of elements (therefore information) described by attributes formed by a common vocabulary. These attributes provide a better search and exploration of information. Faceted search has allowed the creation of new portals, whose goal is to facilitate as good as possible the search of its contents. An evolution of these portals lead to the "exploratory portals" object of the present work; in these advanced versions of portals, users will be guided, and in a sense tempted, in learning. This is due to the presence of two basic components: o The “widget”, in which you can see the attributes ("facets") of various information, with all the values that these attributes can take. o The “canvas”, those are different modes of visualization of the dataset of the elements chosen through the selection of attributes (for example, their geographical distribution). In this context, the main goal of this thesis is to improve the visualization of data and the results of the "explorations". More specifically, this thesis focuses on the visualization of the "facets". The facets can be displayed in different possible ways (called "views”); Currently there are three types of views: one showing the data using an absolute number, one that uses percentages and one that shows the data through written words with different size (larger or smaller), depending on the importance of the attribute. The following improvements were made: 1) Optimization of existing code, to enable the creation of separate various views of the facets, and to automate the insertion of new views. 2) Creation of five new "views": progressiveCircle, chartPie, chartColumn, treeMap and Index. 3) Selection of views to display: the obliged display of too much views, could create a heavy reading of data; moreover, not all views are suitable for all types of data, then, it is inserted the possibility to decide which views show and in which order. 4) Improved communication of the portal: through various expedients, such as adding tooltips or titles. Technical improvements have been tested in the creation of a real case study: the creation of an exploratory portal for papers presented at the conference "Museums and the Web" (www.archimuse.com/conferences/mw.html). This conference is about the contribution of technology to the communication of cultural heritage; it has been active since 1997 and currently includes about 800 articles. The work consisted in the creation of an exploratory portal with the aim of supporting the exploration of the papers presented at the conference according to various attributes: year, author, category, type of institution involved, technologies used, etc.. To make the scenario as realistic as possible, approximately 200 papers have been entered and analyzed. The realization of the case study has shown that the technological improvements make this type of portals more communicative.
“As far as public accessibility is concerned, one has to wonder what the point is of putting more stuff online when we do such a bad job of helping people to explore what is already there” (Norris, 2010). La citazione riassume uno dei problemi che si prospetta quando si affronta la ricerca delle informazioni all’interno della vastità del web. Il più conosciuto tra i metodi di ricerca è senza dubbio quello per parole chiave. Tale metodo presenta limitazioni come, ad esempio, la difficoltà nella formulazione delle “query” di ricerca, le problematiche riguardanti l’impiego di vocabolari diversi ma sovrapposti, l’incapacità di formulare richieste terminologicamente consistenti per domini di conoscenza non noti. Per superare questi ostacoli sono stati creati nuovi e più efficienti sistemi di classificazione e di ricerca delle informazioni. Tra questi sistemi si sta sempre più affermando la ricerca a “faccette” (dall’inglese “facets”: “aspetti”) che dà la possibilità di creare un dataset di elementi (quindi di informazioni) descritti da attributi costituiti da un vocabolario comune. Tali attributi consentono una migliore ricerca ed esplorazione delle informazioni. La ricerca a faccette ha permesso la creazione di nuovi portali che hanno come obiettivo quello di facilitare quanto possibile la ricerca dei propri contenuti. Una evoluzione di questi portali conduce alle “interfacce esplorative” oggetto del presente lavoro; in queste versioni avanzate dei portali, gli utenti saranno guidati, e in un certo senso invogliati, all’apprendimento. Ciò è dovuto alla presenza di due componenti fondamentali: o I “widget” ovvero “specchietti” in cui si possono osservare gli attributi (“facets”) delle varie informazioni, con tutti i valori che tali attributi possono assumere. o I “canvas” ovvero modalità diverse di visualizzazione del dataset degli elementi scelti attraverso la selezione degli attributi (ad esempio, la loro distribuzione geografica). In questo contesto, l’obiettivo primario di questa tesi è il miglioramento della visualizzazione dei dati e dei risultati delle “esplorazioni”. Nello specifico, questa tesi si concentra sulla visualizzazione delle “facets”. Le facets possono essere visualizzate in diversi modi possibili (chiamati “views” - viste); Attualmente esistono tre tipologie di viste: una che mostra i dati servendosi di un numero assoluto, una che utilizza percentuali e una che evidenzia i dati tramite scritte di dimensioni più o meno grandi, in base all’importanza dell’attributo. Sono state apportate le seguenti migliorie: 1) Ottimizzazione del codice già esistente: per consentire la creazione separata delle varie viste delle facets, e automazione dell’inserimento delle varie viste. 2) Creazioni di cinque nuove ”viste”: progressiveCircle, chartPie, chartColumn, treeMap e Index. 3) Scelta delle viste da mostrare: la visualizzazione obbligata di troppe viste potrebbe creare una lettura pesante dei dati; inoltre non tutte le viste sono adatte per tutti i tipi di dati, quindi si è inserita la possibilità di far decidere quali viste mostrare e in che ordine. 4) Miglioramento nella comunicazione del portale: attraverso vari accorgimenti, come l’aggiunta di tooltip o di titoli. Le migliorie tecniche sono state adottate nella realizzazione di un caso di studio reale: la creazione di un portale esplorativo per gli articoli presentati alla conferenza “Museums and the Web”(www.archimuse.com/conferences/mw.html). Questa conferenza tratta dell’apporto delle tecnologie alla comunicazione dei beni culturali; è attiva dal 1997 e a oggi raccoglie all’incirca 800 articoli. Il lavoro è consistito nella creazione di un portale esplorativo con l’obiettivo di supportare l’esplorazione degli articoli presentati alla conferenza secondo vari attributi: anno, autore, categoria, tipologia d’istituzione coinvolta, tecnologie usate, etc. Per rendere lo scenario più reale possibile, si sono inseriti e analizzati all’incirca 200 articoli. La realizzazione del caso di studio ha mostrato che le migliorie tecnologiche rendono questo tipo di portali più comunicativi.
Visualization strategies for an exploratory portal : the case of the museum & the web conference
MARIANI, ALICE MARIA
2012/2013
Abstract
“As far as public accessibility is concerned, one has to wonder what the point is of putting more stuff online when we do such a bad job of helping people to explore what is already there” (Norris, 2010). The quote sums up one of the problems that arises when dealing with the information search within the vastness of the web. The best known of research methods is undoubtedly the one “by keyword”. This method has limitations such as: the difficulty in the formulation of the "query" research, the problems concerning the use of different but overlapping vocabularies, the inability to formulate requests terminologically strong for knowledge domains not known. To overcome these obstacles have created new and more efficient systems of classification and search of information. Among these systems faceted search (from "facets": "aspects") is increasingly asserting and gives the possibility to create a data set of elements (therefore information) described by attributes formed by a common vocabulary. These attributes provide a better search and exploration of information. Faceted search has allowed the creation of new portals, whose goal is to facilitate as good as possible the search of its contents. An evolution of these portals lead to the "exploratory portals" object of the present work; in these advanced versions of portals, users will be guided, and in a sense tempted, in learning. This is due to the presence of two basic components: o The “widget”, in which you can see the attributes ("facets") of various information, with all the values that these attributes can take. o The “canvas”, those are different modes of visualization of the dataset of the elements chosen through the selection of attributes (for example, their geographical distribution). In this context, the main goal of this thesis is to improve the visualization of data and the results of the "explorations". More specifically, this thesis focuses on the visualization of the "facets". The facets can be displayed in different possible ways (called "views”); Currently there are three types of views: one showing the data using an absolute number, one that uses percentages and one that shows the data through written words with different size (larger or smaller), depending on the importance of the attribute. The following improvements were made: 1) Optimization of existing code, to enable the creation of separate various views of the facets, and to automate the insertion of new views. 2) Creation of five new "views": progressiveCircle, chartPie, chartColumn, treeMap and Index. 3) Selection of views to display: the obliged display of too much views, could create a heavy reading of data; moreover, not all views are suitable for all types of data, then, it is inserted the possibility to decide which views show and in which order. 4) Improved communication of the portal: through various expedients, such as adding tooltips or titles. Technical improvements have been tested in the creation of a real case study: the creation of an exploratory portal for papers presented at the conference "Museums and the Web" (www.archimuse.com/conferences/mw.html). This conference is about the contribution of technology to the communication of cultural heritage; it has been active since 1997 and currently includes about 800 articles. The work consisted in the creation of an exploratory portal with the aim of supporting the exploration of the papers presented at the conference according to various attributes: year, author, category, type of institution involved, technologies used, etc.. To make the scenario as realistic as possible, approximately 200 papers have been entered and analyzed. The realization of the case study has shown that the technological improvements make this type of portals more communicative.File | Dimensione | Formato | |
---|---|---|---|
2013_04_Mariani.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Testo della tesi
Dimensione
5.28 MB
Formato
Adobe PDF
|
5.28 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/75043