010false100true score desc 2gaptrue5mapcontentxmltrue50object_type_i,object_type_i_lookup,coverage_period_mt,geographic_area_mt,geographic_coordinates_mt,author_role_mt,contributor_role_mt,org_id_mt,org_role_mt,supervisor_mt,supervisor_id_mi,supervisor_id_mi_lookup,fields_of_research_mi,fields_of_research_mi_lookup,display_type_i,display_type_i_lookup,seo_code_mi,seo_code_mi_lookup,copyright_i,license_i,license_i_lookup,oa_compliance_t,oa_notes_t,grant_id_t,funding_body_t,description_of_resource_t,software_required_t,project_description_t,keywords_mt,project_name_t,project_id_t,isdatasetof_mt,isdatasetof_mt_lookup,notes_t,date_dt,xsd_display_option_mi,xsd_display_option_mi_lookup,file_downloads_i,created_date_dt,updated_date_dt,research_program_mt,title_t,depositor_i,isderivationof_mt,assigned_user_id_mt,assigned_group_id_mi,assigned_group_id_mi_lookup,isdatacomponentof_mt,isannotationof_mt,author_id_mi,author_id_mi_lookup,alternative_title_mt,pid_t,publisher_t,author_mt,contributor_mt,contributor_id_mi,contributor_id_mi_lookup,refereed_i,series_t,journal_name_t,newspaper_t,conference_name_t,book_title_t,identifier_mt,edition_t,subject_mi,subject_mi_lookup,place_of_publication_t,start_page_t,end_page_t,chapter_number_t,issue_number_t,volume_number_t,conference_dates_t,conference_location_t,patent_number_t,country_of_issue_t,description_t,date_available_dt,language_mt,phonetic_title_t,language_of_title_mt,translated_title_t,phonetic_journal_name_t,translated_journal_name_t,phonetic_book_title_t,translated_book_title_t,phonetic_newspaper_t,file_attachment_name_mt,translated_newspaper_t,phonetic_conference_name_t,translated_conference_name_t,issn_mt,isbn_mt,isi_loc_t,prn_t,output_availability_t,na_explanation_t,sensitivity_explanation_t,file_attachment_content_mt,org_unit_name_t,org_name_t,report_number_t,sequence_i,genre_t,genre_type_t,formatted_title_t,formatted_abstract_t,parent_publication_t,convener_t,ismemberof_mt,ismemberof_mt_lookup,link_mt,link_description_mt,rights_t,views_i,scopus_id_t,thomson_citation_count_i,gs_citation_count_i,gs_cited_by_link_t,scopus_citation_count_i,status_i,status_i_lookup,first_author_in_document_derived_t,first_author_in_fez_derived_t,ands_collection_type_t,start_date_dt,end_date_dt,access_conditions_t,extent_t,contact_details_email_mt,contact_details_physical_mt,loc_subject_heading_mt,depositor_affiliation_i,surrounding_features_mt,condition_mt,style_mt,period_mt,category_mt,subcategory_mt,structural_systems_mt,adt_id_t,subtype_t,language_of_parent_title_t,proceedings_title_t,file_description_mt,herdc_code_i,herdc_code_i_lookup,herdc_status_i,herdc_status_i_lookup,institutional_status_i,institutional_status_i_lookup,herdc_notes_t,follow_up_flags_i,follow_up_flags_i_lookup,follow_up_flags_imu_i,follow_up_flags_imu_i_lookup,scopus_doc_type_t,scopus_doc_type_t_lookup,wok_doc_type_t,wok_doc_type_t_lookup,conference_id_i,total_chapters_t,publisher_id_i,translated_proceedings_title_t,native_script_title_t,roman_script_title_t,native_script_book_title_t,roman_script_book_title_t,native_script_journal_name_t,roman_script_journal_name_t,native_script_conference_name_t,roman_script_conference_name_t,total_pages_t,native_script_proceedings_title_t,roman_script_proceedings_title_t,language_of_book_title_mt,language_of_journal_name_mt,language_of_proceedings_title_mt,doi_t,author_count_t,collection_year_dt,location_mt,building_materials_mt,architectural_features_mt,interior_features_mt,sherpa_colour_t,ain_detail_t,rj_2010_rank_t,rj_2010_title_t,rj_2012_rank_t,rj_2012_title_t,rc_2010_rank_t,rc_2010_title_t,herdc_code_description_t,score,citation_t1true60 (sources of information AND ismemberof_mt:bibliuned\:Setopenaire AND display_type_i:198 AND status_i:(2)) 6display_type_idisplay_type_i_lookup_exactkeywords_mftdate_year_tauthor_id_miauthor_id_mi_lookup_exactauthor_mftjournal_name_t_ftsubject_misubject_mi_lookup_exactgenre_type_t_ftismemberof_mftismemberof_mt_lookup_exactsubtype_t_ftscopus_doc_type_t_ftscopus_doc_type_t_lookup_exact(_authlister_t:(1)) AND (status_i:(2)) 31982020-09-01T00:00:00Z2862023-01-11T07:19:29Z2023-01-11T07:19:29ZAutomatic Generation of Domain Knowledge Graph for Recommendation Systems using Open Source Resourcesbibliuned:master-ETSInformatica-LSI-AsanzA group of state-of-the-art recommendation algorithms using Knowledge Graphs are RippleNet [1] or KGAT [2]. However, the main bottleneck to benchmark and work with these algorithms is that they used Microsoft Satori 1 or Freebase 2 (now Google Knowledge Graph 3) as their KG. As Satori and Google Knowledge Graph are commercial KGs and not open source, it's not possible to replicate the results found in the papers on new data, or use these solutions in real applications without paying for access. This limitation is critical in the development of this area of RS. There are researchers working on algorithms and applications, whose results cannot be replicated openly by the science community, hence, going against the scientic method. As researchers, one of our main goals is to make science accessible and replicable for all the scientic community, empowering the development of new knowledge areas. To ll this gap, this Thesis presents a system to generate domain adapted Knowledge Graphs using open source information from Wikidata. These Knowledge Graphs can be used in hybrid Recommendation Systems, that use linked knowledge on the items as side information, combining both Collaborative Filtering and Content Base Filtering strategies. The results show that the proposed system is able to create domain adapted Knowledge Graphs from open source information for recommendation datasets, and that the KGs generated are able to compete with their commercial versions. We have shown that our proposed system creates smaller KGs that are more domain adapted, that have a similar eciency in downstream tasks of Recommendation Systems than commercial KGs. This could be specially relevant in systems that require faster computational times that can be achieved with smaller KGs, as real-time systems, or to save computational cost in high-scale systems. The system ca be used as a reference to evaluate the state-of-the-art algorithms in future works in the area.0Doctoral Thesis1682<a class="citation_author_name" title="Navegar por nombre de Autor de Sanz Olive, Almudena" href="/fez/list/author/Sanz Olive, Almudena/">Sanz Olive, Almudena</a>. (<span class="citation_date">2020</span>). <i><a class="citation_title" title="Click para ver : Automatic Generation of Domain Knowledge Graph for Recommendation Systems using Open Source Resources" href="/fez/view/bibliuned:master-ETSInformatica-LSI-Asanz">Automatic Generation of Domain Knowledge Graph for Recommendation Systems using Open Source Resources</a></i> Master Thesis, <span class="citation_publisher">Universidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Lenguajes y Sistemas Informáticos</span>Recordmaster TesisPublishedIngeniería AeronáuticaUniversidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Lenguajes y Sistemas InformáticosSanz Olive, AlmudenaAlbornoz Cuadrado, Jorge Carrillo deGonzalez-Fierro, Miguelbibliuned:master-ETSInformatica-LSI-Asanzhttp://e-spacio.uned.es/fez/view/bibliuned:master-ETSInformatica-LSI-AsanzengSanzOlive_Almudena_TFM.pdfpresmd_SanzOlive_Almudena_TFM.xmlbibliuned:master-ETSInformatica-LSIbibliuned:Settrabajosfindemasterbibliuned:SetopenaireMáster Universitario en Lenguajes y Sistemas Informáticos (UNED)Set de items trabajo fin de másterSet de openaireSanz OliveAcceso abierto2.479924231982023-01-01T00:00:00Z1392023-12-08T09:43:21Z2023-12-08T09:43:21ZExploring the Power of Large Language Models: News Intention Detection using Adaptive Learning Promptingbibliuned:master-ETSInformatica-TL-AcaballeroIn today’s interconnected global landscape, political news wields significant influence in shaping voter perceptions. Political news stands as a primary source of information for citizens in democratic societies. Understanding the intentions underlying news articles is indispensable for ensuring transparency in political discourse. It empowers the public to make informed decisions and to hold media platforms accountable for the information they disseminate, whether directly or indirectly. Precise identification of news intentions holds the potential to safeguard democratic processes, forming the core focus of this Master’s thesis. Accurate identification of news intentions assumes a central role in Natural Language Processing (NLP) for effective information analysis and categorization. This research project delves into the innovative concept of “Adaptive Learning Prompting”. It harnesses the formidable capabilities of Large Language Models (LLMs) for efficient news intention classification through iterative prompt engineering, with a specific emphasis on resource-constrained scenarios. Employing a dynamic evaluation methodology conducted in iterative batches, this research meticulously tracks model performance. Through trend analysis, the evolution of the proposed model architecture emerges as statistically significant, effectively demonstrating the adaptability inherent in this proposed approach. In summary, this Master’s thesis serves as a testament to the potential of LLMs in adaptive news intention identification. Particularly noteworthy is the “Adaptive Learning Prompting” strategy. While specific performance metrics may exhibit variations, the adaptive approach underscores the feasibility of iterative feedback prompting for operational enhancements, especially in resource-scarce scenarios. These findings transcend the realm of news intention analysis, offering broader applications in NLP, while shedding light on the reasoning capabilities intrinsic to LLMs, thereby amplifying their utility.0Doctoral Thesis3562<a class="citation_author_name" title="Navegar por nombre de Autor de Caballero Hinojosa, Alberto" href="/fez/list/author/Caballero Hinojosa, Alberto/">Caballero Hinojosa, Alberto</a>. (<span class="citation_date">2023</span>). <i><a class="citation_title" title="Click para ver : Exploring the Power of Large Language Models: News Intention Detection using Adaptive Learning Prompting" href="/fez/view/bibliuned:master-ETSInformatica-TL-Acaballero">Exploring the Power of Large Language Models: News Intention Detection using Adaptive Learning Prompting</a></i> Master Thesis, <span class="citation_publisher">Universidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Lenguajes y Sistemas Informáticos</span>Recordmaster TesisPublishedIngeniería InformáticaUniversidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Lenguajes y Sistemas InformáticosCaballero Hinojosa, AlbertoRodrigo Yuste, ÁlvaroCenteno Sánchez, Robertobibliuned:master-ETSInformatica-TL-Acaballerohttp://e-spacio.uned.es/fez/view/bibliuned:master-ETSInformatica-TL-AcaballeroengCaballero_Alberto_TFM.pdfpresmd_Caballero_Alberto_TFM.xmlbibliuned:master-ETSInformatica-TLbibliuned:Settrabajosfindemasterbibliuned:SetopenaireMáster Universitario en Tecnologías del Lenguaje (UNED)Set de items trabajo fin de másterSet de openaireCaballero HinojosaAcceso abierto2.296198131982021-07-01T00:00:00Z4572021-10-29T19:37:18Z2021-10-29T19:37:18ZFake News Detection using news content and user engagementbibliuned:master-ETSInformatica-ICD-MperezFake news are purposefully designed to be misleading, and their success depends mostly on their readers [ ¨OG17]. Due to the nature of social networks, fake news can be quickly propagated, potentially causing a great damage to the society. Moreover, sociological phenomena like echo chambers or polarization [Sil+16], and psychological factors like confirmation bias [Del+16] or overconfidence to be fooled by fake news 1, create the perfect playground for misinformation [Tac +17] [Del+18]. Platforms have recently adopted measures 2 to discourage users sharing content without reading it first, but these measures are still not fully enforced and easy to bypass. An automatic fake news detection system that blocks or warns users about possibly misleading information will be needed in the near future, espe-cially with the high volume of information that is shared through these websites. Several models have been proposed to detect fake news by analyzing linguistic features [HA17] [Pot +17][BS19] [KGN21], but these are often not enough [Shu+18] to distinguish fake news from real ones. Research is now focusing on including user engagement information to existing contentbased models [RSL17] [Del+18] [SML19]. Some systems have been proposed that only use information from these engagements [Tac+17]. Part of the research has focused on creating training datasets [SW18] [Shu+18] for this task, crawling news from fact-checking websites and fetching user engagements using the public APIs offered by social networks. In this work, we develop and test different fake news detection systems using information from news articles and user engagements in social networks. Two different architectures are used. The main one is based on Deep Learning, and can process news content and user engagements. The second one is based on well-known algorithms like logistic regression, SVM, random forest, LightGBM or XGBoost; it can only process news content and is used as a performance baseline for our Deep Learning models. We use the FakeNewsNet dataset [Shu+18], which contains real and fake news from two factchecking sources. For each news piece, this dataset contains the scraped news article, as well as tweets and retweets related to each news, and user profiles of the users involved in these tweet, including the user’s timeline, followers and followees, although not all this information will be used. Our work starts with an exploratory data analysis on the train set, where we highlight the main charac-teristics of the dataset. Then, we carry out a series of tests on both architectures, taking news from each set of news, with different subsets of features and with various textual representation techniques. Additionally, we perform an ablation test on the Deep Learning architecture, to understand how individual features behave and how do they complement each other. Our results clearly show that our architecture is able to capture much information from user engagements, and that including user interactions gives better results than models using only information from news articles. With this work, our main contribution is a Deep Learning architecture capable of handling varying-length sequences of engagements for each piece of news, while also extracting all the information from them without padding or truncating to fixed-size sequences. We take advantage of recent innovations in frameworks like Tensorflow to process non-tabular-shaped data, which allows to directly include unaggregated features, minimizing the preprocessing required before the input data is fed to the model. This architecture can perform complex summarizations, such as a trainable recurrent layer that takes a sequence of user engagements in the same order as they were published, and outputs a vector that summarizes the whole user engagement sequence0Doctoral Thesis3812<a class="citation_author_name" title="Navegar por nombre de Autor de Pérez Madre, Mario" href="/fez/list/author/Pérez Madre, Mario/">Pérez Madre, Mario</a>. (<span class="citation_date">2021</span>). <i><a class="citation_title" title="Click para ver : Fake News Detection using news content and user engagement" href="/fez/view/bibliuned:master-ETSInformatica-ICD-Mperez">Fake News Detection using news content and user engagement</a></i> Master Thesis, <span class="citation_publisher">Universidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Inteligencia Artificial</span>Recordmaster TesisPublishedIngeniería InformáticaUniversidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Inteligencia ArtificialPérez Madre, MarioRodrigo Yuste, Álvarobibliuned:master-ETSInformatica-ICD-Mperezhttp://e-spacio.uned.es/fez/view/bibliuned:master-ETSInformatica-ICD-MperezengPerezMadre_MarioTFM.pdfpresmd_PerezMadre_MarioTFM.xmlbibliuned:master-ETSInformatica-ICDbibliuned:Settrabajosfindemasterbibliuned:SetopenaireMáster Universitario en Ingeniería y Ciencia de Datos (UNED)Set de items trabajo fin de másterSet de openairePérez MadreAcceso abierto2.283340531982020-10-02T00:00:00Z1412021-09-23T20:33:14Z2021-09-23T20:33:14ZOptimización de árboles de estrategia unicriterio y de coste-efectividadbibliuned:master-ETSInformatica-IAA-AmgilIntroduction: A cost-effectiveness analysis (CEA) helps us select the most effective intervention for the financial budget we have. In these analyzes through CEP, we generated several lambda intervals, each with the optimal intervention, its cost and expected effectiveness. Being able to represent these lambda intervals, through a graphic model, will help us see all the available information and will speed us up to make a decision. If we can optimize this tree, it will be easier to study and understand it. Objective: To find the algorithm that returns the best optimization of the generated tree after applying the deterministic CEA analysis. Methods: Applying pruning techniques on the tree nodes (variable exchange, elimination of redundancies, lambda displacement) in a search algorithm, we will be able to get partial results of the search optimization, until we reach the most optimal tree optimization result created with CEP results. Results: Once optimized with the most appropriate pruning techniques and algorithms, we will have a graphic model where the different options can be observed in a more effective and clear way than when we use a textual description. Conclusion: Through the open source software tool "OpenMarkov" the possibility of graphically displaying the result of a deterministic CEA analysis has been implemented. With this new tool we can graphically evaluate the results that were previously shown in a table, and that did not allow us to appreciate in the same detail, each of the concepts that are observed in each branch of the tree.0Doctoral Thesis2772<a class="citation_author_name" title="Navegar por nombre de Autor de Gil González, Ángel Miguel" href="/fez/list/author/Gil González, Ángel Miguel/">Gil González, Ángel Miguel</a>. (<span class="citation_date">2020</span>). <i><a class="citation_title" title="Click para ver : Optimización de árboles de estrategia unicriterio y de coste-efectividad" href="/fez/view/bibliuned:master-ETSInformatica-IAA-Amgil">Optimización de árboles de estrategia unicriterio y de coste-efectividad</a></i> Master Thesis, <span class="citation_publisher">Universidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Inteligencia Artificial</span>Recordmaster TesisPublishedIngeniería InformáticaUniversidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Inteligencia ArtificialGil González, Ángel MiguelArias Calleja, ManuelLuque Gallego, ManuelDíez Vegas, Francisco Javierbibliuned:master-ETSInformatica-IAA-Amgilhttp://e-spacio.uned.es/fez/view/bibliuned:master-ETSInformatica-IAA-AmgilspaGil_Gonzalez_Angel_Miguel_TFM.pdfpresmd_Gil_Gonzalez_Angel_Miguel_TFM.xmlbibliuned:master-ETSInformatica-IAAbibliuned:Settrabajosfindemasterbibliuned:SetopenaireMáster Universitario en I.A. Avanzada: Fundamentos, Métodos y Aplicaciones (UNED)Set de items trabajo fin de másterSet de openaireGil GonzálezAcceso abierto2.161664244324444