Submitted by JÚLIO HEBER SILVA (julioheber@yahoo.com.br) on 2016-11-29T17:24:41Z No. of bitstreams: 2 Dissertação - Patrícia Araújo Vieira - 2016.pdf: 1738341 bytes, checksum: a398b5fde24860c1899568800dab505b (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Jaqueline Silva (jtas29@gmail.com) on 2016-11-30T15:38:17Z (GMT) No. of bitstreams: 2 Dissertação - Patrícia Araújo Vieira - 2016.pdf: 1738341 bytes, checksum: a398b5fde24860c1899568800dab505b (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2016-11-30T15:38:17Z (GMT). No. of bitstreams: 2 Dissertação - Patrícia Araújo Vieira - 2016.pdf: 1738341 bytes, checksum: a398b5fde24860c1899568800dab505b (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2016-09-06 / Innovation in the public sector, and in particular technological innovation, is an example of change within public organizations. One form of technological innovation is e-government, which is defined as the modernization of public administration through the usage of information and communication technologies (ICTs). Examples that stands out of the e-government programs include the ones developed by the tax administration, such as the Electronic Invoice (NF-e). Although it is an important program and used throughout Brazil, there is still little study about it and there is no evaluation of the performance of this type of public fiscal policy in the country’s Treasury Departments. The aim of this study was to evaluate, according to the parameters of effectiveness and effectiveness itself, the Electronic Invoice program in the State of Goiás, using the guidelines of the GESPÚBLICA guide for evaluating public policies. Specifically, it was intended to make a diagnosis to identify the main problems with NF-e in the State, to propose solutions and implement those that were feasible during the execution of this research. In order to measure the effectiveness, a SEFAZ database was used to analyze the sharing of tax documents. Furthermore, to measure the effectiveness, a questionnaire was applied to the fiscal auditors of the State’s Department of Finance about the changes in the inspection after the implantation of NF-e. Through the evaluation of public policy, it was identified that in relation to effectiveness, the NF-e program in Goiás does not match the goals proposed for synchronization of fiscal documents. It was also found that the tax of sending the fiscal documents to the National Environment does not reach all the authorized documents and the period of time to send and to receive the fiscal documents is very long. Regarding effectiveness, the conclusion is that that the NF-e program has brought improvements to the inspection and audit processes to the State’s Department of Finance. The main improvement was the speed of the audits, due to the increase of the facility and the velocity in obtaining the data for inspection. It is also worth mentioning the real-time monitoring of the operations, which reduces the response time of audits in relation to possible fraud and evasion. Regarding the work intervention, changes were made in the synchronism system of fiscal documents, according to the solution proposed from the diagnosis made and as a result a significant improvement in the timing of the fiscal documents to the National Environment was obtained. / A inovação no setor público, e em especial a inovação tecnológica, é um exemplo de mudança dentro das organizações públicas. Uma forma de inovação tecnológica é o governo eletrônico, que é definido como a modernização da administração pública, por meio do uso de tecnologias de informação e comunicação (TICs). Como exemplos de programa de governo eletrônico, se destacam os programas elaborados pela administração tributária, como a Nota Fiscal Eletrônica (NF-e). Apesar de ser um programa importante e utilizado em todo Brasil, ainda há pouco estudo sobre o mesmo e não há avaliação sobre o desempenho deste tipo de política pública fiscal nas Secretarias de Fazenda do país. O objetivo desse trabalho foi avaliar, em relação aos parâmetros de eficácia e efetividade, o programa da Nota Fiscal Eletrônica em Goiás, utilizando as diretrizes do guia da GESPÚBLICA de avaliação de políticas públicas. De forma especi- fica, pretendeu-se realizar um diagnóstico para identificar os principais problemas com a NF-e no Estado de Goiás, propor soluções e implementar aquelas que fossem viáveis durante a execução desta pesquisa. Para medir a eficácia utilizou-se coleta na base de dados da SEFAZ, com a finalidade de analisar o compartilhamento de documentos fiscais e para medir a efetividade foi aplicado questionário aos auditores fiscais da Secretaria da Fazenda do Estado de Goiás, sobre as alterações na fiscalização após a implantação da NF-e. Por meio da avaliação da política pública, identificou-se que em relação a eficácia, o programa NF-e em Goiás não está de acordo com as metas propostas para sincronização dos documentos fiscais. Constatou-se que a taxa de envio dos documentos fiscais para o Ambiente Nacional não chega à totalidade dos documentos autorizados e o tempo de envio e recebimento dos documentos fiscais é alto. No tocante à efetividade pode-se concluir que o programa NF-e trouxe melhorias para os processos de fiscalização e auditoria da Secretaria da Fazenda de Goiás. A principal melhoria apontada foi a agilidade nas auditorias, devido a maior facilidade e celeridade na obtenção dos dados para a fiscalização. Destaca-se também o acompanhamento em tempo real das operações, que diminuiu o tempo de resposta nas auditorias em relação à possíveis fraudes e sonegações. Sobre a intervenção do trabalho, foram realizadas alterações no sistema de sincronismo de documentos fiscais, conforme solução proposta a partir do diagnóstico realizado e como resultado foi obtido uma melhora significativa no tempo de sincronismo dos documentos fiscais ao Ambiente Nacional.

This master's thesis desribes the import process of aggregate invoices and prepayments into the gas trader's SAP IS-U information system. It explains some of the essential SAP terms and contains an analysis of final product requirements. Additionally, it concerns the implementation of the designed solution itself and describes achieved results. In conclusion, it mentions possible extensions.

This diploma thesis deals with claims in selected capital company. Specifically, it fosuces on the trade receivables. Through the results of analysis of current receivables, proposes measure that can help improve the situation in he selected company.

Companies often process invoices manually, therefore automation could reduce manual labor. The aim of this thesis is to evaluate which OCR-engine, Tesseract or OCRopus, performs best at interpreting invoices. This thesis also evaluates if it is possible to use machine learning to automatically process invoices based on previously stored data. By interpreting invoices with the OCR-engines, it results in the output text having few spelling errors. However, the invoice structure is lost, making it impossible to interpret the corresponding fields. If Naïve Bayes is chosen as the algorithm for machine learning, the prototype can correctly classify recurring invoice lines after a set of data has been processed. The conclusion is, neither of the two OCR-engines can interpret the invoices to plain text making it understandable. Machine learning with Naïve Bayes works on invoices if there is enough previously processed data. The findings in this thesis concludes that machine learning and OCR can be utilized to automatize manual labor. / Företag behandlar oftast fakturor manuellt och en automatisering skulle kunna minska fysiskt arbete. Målet med examensarbetet var att undersöka vilken av OCR-läsarna, Tesseract och OCRopus som fungerar bäst på att tolka en inskannad faktura. Även undersöka om det är möjligt med maskininlärning att automatiskt behandla fakturor utifrån tidigare sparad data. Genom att tolka text med hjälp av OCR-läsarna visade resultaten att den producerade texten blev språkligt korrekt, men att strukturen i fakturan inte behölls vilket gjorde det svårt att tolka vilka fält som hör ihop. Naïve Bayes valdes som algoritm till maskininlärningen och resultatet blev en prototyp som korrekt kunde klassificera återkommande fakturarader, efter att en mängd träningsdata var behandlad. Slutsatsen är att ingen av OCR-läsarna kunde tolka fakturor så att resultatet kunde användas vidare, och att maskininlärning med Naïve Bayes fungerar på fakturor om tillräckligt med tidigare behandlad data finns. Utfallet av examensarbetet är att maskininlärning och OCR kan användas för att automatisera fysiskt arbete.

Manual invoice processing is a time-consuming and error prone task which has proven to be done more efficiently by introducing automation software that minimizes the need for human input. Amazon Textract is a software as a service provided by Amazon Web Services for that purpose. It has been developed to extract document data from both general and financial documents, such as receipts and invoices, by using machine learning models. The service is available in multiple widely spoken languages, but not in Swedish as of the time of writing this thesis. This thesis explores the potential and accuracy of Amazon Textract in extracting data from Swedish invoices by using the English setting. Specifically, the accuracy of extracting line items as well as Swedish letters are examined. In addition, the potential of correcting incorrectly extracted data is explored. This is achieved by testing certain defined categories on each invoice by comparing the Amazon Textract extractions with the correct labeled data. These categories include emptiness, meaning no data was extracted, equality, missing and added line items, as well as missing and added characters that have been added or removed from otherwise correct line item strings. The invoices themselves are divided into two categories, namely structured and semi-structured invoices. The tests are mainly conducted on the service’s dedicated API method for data extraction of financial documents, but a comparison with the table extraction API method is also made to gain more insight in Amazon Textract’s capability.  The results suggest that Amazon Textract is quite inaccurate when extracting line item data from Swedish invoices. Therefore, manual post processing of the data is generally needed to ensure its correctness. However, it showed better results in extracting data from structured invoices, where it scored 70% in equality and 100% in 2 out of 6 invoice layouts. The Swedish character accuracy was 66%.

Buy Now Pay Later (BNPL) is a financial service whereby customers defer payment on a purchase against a short-term debt. While BNPL services have a long history, digital invoice services are now the largest market for BNPL. For the study, two of the largest providers in Sweden, and their checkout interfaces were reviewed for dark patterns. Dark patterns are instances where designers exploit the knowledge of human behaviour (e.g., behavioural psychology) and cognition (e.g., cognitive psychology) in order to coerce the customers into performing target actions. For digital invoice services, the target actions are for the customers to select the credit instalments which profits their business model, as a large percentage of their income are from postponed payments. This study employs three research methods to answer the research questions “What existing dark patterns are identified in digital invoice services, and what are the impacts of dark patterns in the context of digital invoice services?”. First, existing dark patterns were identified through a systematic literature review. Second, the resulting dark patterns were applied to an analysis of the two digital invoice services, where ten dark patterns were identified. Finally, an interface of a digital invoice service was designed, prototyped, and evaluated to test the impact of the identified dark patterns in the context of digital invoice services. Analysis suggests that the users are being coerced through design into the target actions, which were selecting the credit instalments (X2 = 5.84, df=1). The results is being further discussed in means of the potential debt users face from dark patterns in financial interfaces. / Buy Now Pay Later (BNPL) är en finansiell tjänst där användarna skjuter upp betalningar mot en kortfristig skuld. BNPL-tjänster har en lång historia, men digitala faktureringstjänster är nu den största inom marknaden för BNPL. För studien granskades Sveriges två av de största faktureringstjänster och deras användargränssnitt för “Dark Patterns”. Dark Patterns är designmönster som utnyttjar kunskapen om mänskligt beteende (t.ex. beteendepsykologi) och kognition (t.ex. kognitiv psykologi) för att uppmana kunderna att utföra riktade åtgärder. För digitala faktureringstjänster är de riktade åtgärderna att användarna väljer kreditbetalningar vilket uppfyller deras affärsmodell, eftersom en av deras huvudsakliga inkomstkällor är från användare som skjuter upp sina betalningar. Denna studie använder tre forskningsmetoder för att svara på forskningsfrågan ”Vilka befintliga “Dark Patterns” kan identifieras i digitala faktureringstjänster, och vad är påverkan av “Dark Patterns” i samband med digitala faktureringstjänster?”. Först identifierades befintliga “Dark Patterns” genom en systematisk litteraturstudie. Sen tillämpades de identifierade “Dark Patterns” i en analys av digitala faktureringstjänster, där tio “Dark Patterns” identifierades. Slutligen skapades en prototyp vars gränssnitt utvärderas för att testa påverkan av de identifierade “Dark Patterns” i samband med digitala faktureringstjänster. Analys tyder på att användarna uppmanas genom design till riktade åtgärder för de digitala faktureringsstjänsterna som valde kreditbetalningarna (X2 = 5,84, df = 1).

When an organization receives invoices, accountants specify accounts and cost centers related to the purchases. This thesis investigated automated decision support with machine learning that gives suggestions to the accountant of what accounts and cost centers that can be used for invoices. The purpose was to identify benefits and risks of using machine learning automation for invoice processing and evaluate the performance of this technology. It was found that machine learning-based decision support for invoice processing is perceived to be beneficial by saving time, reducing the mental effort, create more coherent bookkeeping, detect errors, and enabling higher levels of automation. However, there are also risks related to implementing automation with machine learning. There is a high variety of how accounts and cost centers are used in different organizations and an uneven performance can be expected due to that some invoices are more complex to process than others. Machine learning experiments were conducted which indicated that the accuracy of suggesting the correct account was 73–76%. For cost centers, the accuracy was 50–62%. A method for filtering machine learning output was developed with the aim of raising the accuracy of the automated suggestions. With this method, the limited amount of suggestions that passed the filter achieved accuracy up to 100%. / När en organisation tar emot fakturor anges konton och kostnadsställen relaterade till inköpen. Detta examensarbete undersökte automatiserat beslutsstöd med maskininlärning som ger förslag på vilka konton och kostnadsställe som kan användas för fakturor. Syftet var att identifiera fördelarna och riskerna med att använda automatisering med maskininlärning för fakturahantering och utvärdera teknikens prestanda. Resultaten visade att maskininlärningsbaserat beslutsstöd för fakturabehandling uppfattas vara fördelaktigt genom att spara tid, minska mentala ansträngning, skapa mer sammanhängande bokföring, upptäcka fel, och möjliggöra högre automatiseringsnivåer. Men det finns också risker relaterade till implementering av automatisering med maskininlärning. Det är en stor variation gällande hur konton och kostnadsställen används i olika organisationer och en ojämn prestanda kan förväntas på grund av att vissa fakturor är mer komplexa att bokföra än andra. Maskininlärningsexperiment genomfördes som indikerade att korrektheten i att föreslå rätt konto var 73–76%. För kostnadsställe var korrektheten 50–62%. En metod för att filtrera maskininlärnings-förslagen utvecklades i syfte att höja korrektheten för automatiseringen. Med denna metod uppnådde den begränsade mängden förslag som passerade filtret en korrekthet upp till 100%.

Forskningsfrågor: Vilka möjligheter och utmaningar finns med användningen av maskininlärning inom faktureringsprocessen? Hur har maskininlärning påverkat felkällorna inom faktureringsprocessen? Syfte:    Denna studie syftar till att beskriva automatisering inom bokföringsprocessen, med fokus på fakturahantering. Detta inkluderar en analys av hur maskininlärning kan tillämpas för att automatisera datainmatning och genomföra felkontroller i bokföring, samt hur väl denna teknologi kan koppla information till korrekt kontoklass. Metod:   I denna studie genomfördes datainsamlingen med hjälp av kvalitativa metoder, specifikt genom användning av semistrukturerade intervjuer. Urvalsgruppen bestod av systemutvecklare, produktutvecklare, IT-experter och en auktoriserad redovisningskonsult.  Slutsats:   Denna studie ger insikt i både möjligheter och utmaningar med maskininlärning inom faktureringsprocessen. Automatiseringen underlättar tolkning av fakturor och identifiering av mönster, vilket frigör tid för personal att fokusera på mer avancerade uppgifter. Fel som upptäcks i efterhand kan vara kostsamma att korrigera och det är kundens ansvar att säkerställa bokföringens korrekthet och vara medveten om potentiella felkällor. Ökad medvetenhet om systemets begränsningar och styrkor kan bidra till ökad trovärdighet och tillförlitlighet i bokföringsprocessen. / Research questions: What opportunities and challenges exist with the use of machine learning in the invoicing process? How has machine learning impacted error sources within the invoicing process? Purpose: This study aims to describe automation within the accounting process, with a focus on invoice management. This includes an analysis of how machine learning can be applied to automate data entry and conduct error checks in accounting, as well as how effectively this technology can link information to the correct account classification. Method: In this study, data collection was conducted using qualitative methods, specifically through the use of semi-structured interviews. The sample group consisted of software developers, product developers, IT experts, and a certified public accountant.   Conclusion: This study provides insights into both the opportunities and challenges of machine learning within the invoicing process. Automation facilitates the interpretation of invoices and the identification of patterns, thus freeing up time for staff to focus on more advanced tasks. Errors detected post hoc can be costly to rectify, and it is the customer's responsibility to ensure the accuracy of accounting and be aware of potential sources of errors. Increased awareness of the system's limitations and strengths can contribute to enhanced credibility and reliability in the accounting process.

財政部於2006年底建置完成的電子發票整合服務平台，提供不同產業間之買賣雙方一個具有公信力的交易稽核平台。企業在此可利用電子發票向銀行進行貸款，完成融資服務，由於貸款跨越企業與銀行，而各個不同組織間，其流程中會有金、商流共同所需之資訊，但是彼此所需的資訊格式的不同，讓整個流程無法一氣喝成。 本研究提出一個以雲端運算為基礎的服務導向架構電子發票加值平台，透過此平台企業可將原本的發票融資，轉換成利用電子發票進行線上融資，再透過XML-based訊息轉換器將企業端電子發票XML格式轉換為處理帳務之XBRL格式，以及銀行端之金流訊息FXML格式，並利用雲端運算服務作為資料儲存與呈現的基礎，而服務導向架構提供了完成此跨組織金、商流活動所需的平台一個良好架構。 在電子發票加值平台中，處理電子發票加值服務時會運用發票上的金流與商流資訊，基於安全性的考量，所以不能將資料都儲存於雲端運算的資源中，本研究的資料中心利用分散式資料儲存方式，將機密的資料儲存於企業端，減少企業使用服務的疑慮，並透過分散式資料擷取/儲存機制對不同的資料庫存取服務所需之資料，讓企業可以更放心且便利的使用服務。 / Ministry of Finance, R.O.C builds a reliable E-Invoice platform in 2006, to provide buyers and sellers a credible audit platform. Through E-Invoice platform, companies could complete loan service. Because loan service crosses the banks and enterprises, they will need the same information in loan process. However the information to each other is different formats, so that the whole process can not complete straight through. This study proposed a Service-oriented architecture E-Invoice value-added platform which based on cloud computing services. Through this platform, companies can change invoice loan service into E-Invoice on-line loan service. In this platform, they not only can use a XML-based message converter to convert business XML format to financial information XBRL format and cash flow information FXML format but also can use cloud computing services to store and present data. However Service-oriented architecture can provide this cross-organizational activity a suitable architecture. In E-Invoice value-added platform, E-Invoice on-line loan service will handle the information of cash flow and business flow, but based on security considerations, we can not stored all the data in the cloud. However this study use decentralized data center to store confidential information in the enterprise client to solve this problem and use some mechanism to extract/store data from different data center. Therefore Companies can reduce their concerns of using services and use services conveniently.

目前，因應不同需求與應用情境的各式電子發票系統已被建立。儘管電子發票系統、區塊鏈技術與其相關的文件及平台已經存在，許多研究學者及廠商仍希望推展此技術並應用於其它領域。 本研究的目的在以區塊鏈技術為基礎建構電子發票平台，希望提出一種區塊鏈技術的新應用情境。首先，透過資訊計量學的方法分析980篇電子發票及區塊鏈的相關文獻以了解電子發票及區塊鏈的發展；接著，探討比特幣(Bitcoin)與區塊鏈(Blockchain)間的重要觀念與技術；最後，利用區塊鏈技術中的一種以太坊(Ethereum)來建立電子發票平台。 本研究中六項建構電子發票系統的需求如資料完整性、可驗證性、可用性、可稽核性、不可否認性與可溯源性可被達成。此架構不僅適用於發行一般的消費性電子發票更適用於向企業提供短期資金借貸如發票融資的應用情境。未來，在不同的實作需求下，將可探討整合區塊鏈另一項技術開放鏈(Openchain)以建立混合公開與私有節點的應用。 / Different e-invoice systems are designed to meet different needs and application scenarios. Although these systems areuseful and readily available, researchers and businesses hope to be able to expand the coverage of e-invoice systems using Blockchain technologies. This study proposes an e-invoice system based on Blockchain technologies which shows anotherapplication scenario of Blockchain technology and e-invoice system. First, a review of about 980 publicationson e-invoicing and Blockchain was carried outwithinformetrics used for further analysis so as to better understandthe development of e-invoice and Blockchain technologies. Second, the concepts and mechanisms of Bitcoin and Blockchain technologies were reviewed. Finally, an e-invoice system based on Ethereumwas proposed. The proposed system meets the six needs of an e-invoice system: integrity, authenticity, availability, auditability, non-repudiation and traceability. It is not only suitable for issuing general invoices but also can play a role in further invoicing applications such as invoice financing as well as agri-food supply chain traceability. In the future, Openchain or hybrid Blockchain infrastructure that combines public and private nodes can be consideredfor use in this proposed e-invoice system.