|
|
Question Answering over Structured Data
Birger, Mark ; Otrusina, Lubomír (oponent) ; Smrž, Pavel (vedoucí práce)
This thesis deals with question answering over structured data. In knowledge databases, a structured data is usually represented by graphs. However, to satisfy information needs using natural language interfaces the system is required to hide the underlying schema from users. A question answering system with a schema-agnostic graph-based approach was developed as a part of this work. In contrast to traditional question answering systems that rely on deep linguistic analysis and statistical methods, the developed system explores provided graph to yield and reuse semantic connection for a known question-answer pair. Lack of large domain-specific structured data made us perform evaluation with the help of prominent open linked datasets such as Wikidata and DBpedia. Quality of separate answering stages and the approach in general was evaluated using adapted evaluation dataset and standard metrics.
|
|
|
Unified Access to Requirements in JIRA
Pindur, Daniel ; Vašíček, Ondřej (oponent) ; Fiedor, Jan (vedoucí práce)
The goal of this work is to create an interface for accessing requirements resources stored in Jira Software with the Requirements for Jira (R4J) plugin using an OSLC adaptor for the Requirements Management specification. The adaptor has been split into two to explore the possibility of not needing the R4J plugin - one for the Jira Software (satisfying all requirements of the Requirement Management specification) and one for the R4J plugin (providing additional functionality). Both adaptors have been implemented by utilizing the Eclipse Lyo tooling and the OSLC4J library. The basic overview of the fundamental technologies of OSLC, OSLC Core and Requirement Management specifications, Jira and R4J is provided. The thesis contains a detailed summary of the adaptors' design, implementation, and testing process, as well as an evaluation of the results and used technologies.
|
|
|
Vizuální browser grafových dat
Stenchlák, Štěpán ; Nečaský, Martin (vedoucí práce) ; Skopal, Tomáš (oponent)
Jeden ze způsobů, jak publikovat data ve strojově čitelné podobě na internetu, je formou grafu. Taková data je pak velmi snadné propojovat mezi různými datovými zdroji a vytvořit tak velkou síť propojených dat. Abychom tato data mohli vizualizovat, musíme používat nástroje na procházení grafu, které mohou být často nepraktické, protože dat je typicky hodně na to, abychom je zobrazili všechny najednou. Cílem této práce je vytvořit webovou aplikaci, která je schopna tato grafová data vizualizovat, poskytovat k nim informace a procházet je pomocí takzvaných konfigurací. Konfigurace popisuje, jaká data z velké množiny dat na internetu chceme zobrazit, jak je vizualizujeme a jak se na ně můžeme dívat. Dovolí nám tedy odfiltrovat stovky neza- jímavých informací a odstíní tak uživatele od složité sítě propojených dat, aby se mohl věnovat pouze těm, která ho zajímají. 1
|
|
|
Decentralized Web-based Data Storage for LinkedPipes Applications using Solid
Orumbayev, Altynbek ; Klímek, Jakub (vedoucí práce) ; Škoda, Petr (oponent)
The following thesis work is focused on developing a decentralized web-based storage solution based on Solid project for the LinkedPipes Applica- tions platform and demonstrate the full benefits of decentralized storage and the Solid technology. The Solid project, started by Sir Tim Berners-Lee, is focused on decentralizing the World Wide Web and decoupling data from applications on the Internet. The results of the practical work satisfy all defined functional and non-functional requirements stated by the LinkedPipes Applications project. The results of the work are integrated into the LinkedPipes Applications platform as a set of several software libraries allowing interaction with Solid Servers. iv
|
|
|
Analysis of the development direction of a conceptual academic library resource sharing service based on a case study of DXY
Niu, Xiaofei ; Zhang, Ling ; Han, Li
Under a global environment characterized by open and complex information, academic library resource-sharing practitioners should observe the information-service market beyond a “library-oriented” setting and rethink the development direction of the service. To investigate this issue, this paper presents an analysis of the development trajectory and characteristics of DXY, a Chinese information service enterprise. This study also discusses the development of academic library resource-sharing service from three dimensions, namely, service targets, service contents, and service roles. Findings show that the academic library resource-sharing practitioners should break through the traditional cognitive framework of “academic users on campus” and “document providers,” and furthermore, consider upgrading the service to assume a comprehensive role as information sharer, information connector, and platform operator simultaneously. This change will enable the move toward a knowledge service orientation.
Plný text:  PDF
|
|
|
Blend and deblend Linked open data in a Consortium
Pallarès, Jordi
Working in a Consortium give us the perspective to see the benefits of blending ideas to create a applications from a central data/point. In some cases we found the Institution want to deblend or „not blend“ with the consortium and prefer or they see more benefits to made his own aplication. We explain our experience in the Consortium blending and not blending desicions and explain two projects in linked open data to show examples of this two ways. One project We blend all the authorities and in other we not blend in the case of Thesaurus of the University of Barcelona using Skos format.
Prezentace: idr-1248_1 - PDF
Videozáznam: ELAG2018-Pallares -  MP4
|
|
|
Question Answering over Structured Data
Birger, Mark ; Otrusina, Lubomír (oponent) ; Smrž, Pavel (vedoucí práce)
This thesis deals with question answering over structured data. In knowledge databases, a structured data is usually represented by graphs. However, to satisfy information needs using natural language interfaces the system is required to hide the underlying schema from users. A question answering system with a schema-agnostic graph-based approach was developed as a part of this work. In contrast to traditional question answering systems that rely on deep linguistic analysis and statistical methods, the developed system explores provided graph to yield and reuse semantic connection for a known question-answer pair. Lack of large domain-specific structured data made us perform evaluation with the help of prominent open linked datasets such as Wikidata and DBpedia. Quality of separate answering stages and the approach in general was evaluated using adapted evaluation dataset and standard metrics.
|
|
|
Historie, současnost a budoucnost výměnných formátů bibliografických dat
Rösslerová, Klára ; Vlasák, Rudolf (vedoucí práce) ; Sklenák, Vilém (oponent) ; Drobíková, Barbora (oponent)
Disertace je zaměřena na výměnné formáty bibliografických dat, které jsou v knihovnické praxi používány od 60. let minulého století. Přes několik desetiletí nedošlo v této oblasti k příliš velké změně, přestože jsou předmětem odborných diskuzí. Od nového tisíciletí dochází k vyššímu tlaku na změnu. Cílem disertace je shrnutí aktivit na tomto poli a nastínění cesty možného vývoje. Nejprve jsou jednotlivé v minulosti i současnosti používané formáty analyzovány. Na základě analýzy jednotlivých současných projektů, ale i dotazníkových šetření, lze směr, kterým se tato oblast bude nyní ubírat, odhadnout. Aktivity nyní směřují k publikačnímu modelu založenému na principu propojených dat (linked data) při zachování současné katalogizační praxe. Klíčová slova: výměnné formáty, bibliografická data, formát MARC, formát MARC 21, propojená data, linked data, BIBFRAME, Schema.org, katalogizace 2
|
host ::
RSS.