Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8846972
Julian Risch, Ralf Krestel
A patent examiner needs domain-specific knowledge to classify a patent application according to its field of invention. Standardized classification schemes help to compare a patent application to previously granted patents and thereby check its novelty. Due to the large volume of patents, automatic patent classification would be highly beneficial to patent offices and other stakeholders in the patent domain. However, a challenge for the automation of this costly manual task is the patent-specific language use. To facilitate this task, we present domain-specific pre-trained word embeddings for the patent domain. We trained our model on a very large dataset of more than 5 million patents to learn the language use in this domain. We evaluated the quality of the resulting embeddings in the context of patent classification. To this end, we propose a deep learning approach based on gated recurrent units for automatic patent classification built on the trained word embeddings. Experiments on a standardized evaluation dataset show that our approach increases average precision for patent classification by 17 percent compared to state-of-the-art approaches.
{"title":"Learning Patent Speak: Investigating Domain-Specific Word Embeddings","authors":"Julian Risch, Ralf Krestel","doi":"10.1109/ICDIM.2018.8846972","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8846972","url":null,"abstract":"A patent examiner needs domain-specific knowledge to classify a patent application according to its field of invention. Standardized classification schemes help to compare a patent application to previously granted patents and thereby check its novelty. Due to the large volume of patents, automatic patent classification would be highly beneficial to patent offices and other stakeholders in the patent domain. However, a challenge for the automation of this costly manual task is the patent-specific language use. To facilitate this task, we present domain-specific pre-trained word embeddings for the patent domain. We trained our model on a very large dataset of more than 5 million patents to learn the language use in this domain. We evaluated the quality of the resulting embeddings in the context of patent classification. To this end, we propose a deep learning approach based on gated recurrent units for automatic patent classification built on the trained word embeddings. Experiments on a standardized evaluation dataset show that our approach increases average precision for patent classification by 17 percent compared to state-of-the-art approaches.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128248886","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8846987
Tze Wei Liew, Su-Mae Tan, Chin Lay Gan
Virtual agents can be integrated in e-learning environments to encourage learning behavior through persuasive and motivational messages. In this study, we aimed to investigate the effects of agent’s message-frame, i.e., gainframe and loss-frame on cognitive load and intrinsic motivation of learners interacting with motivational virtual agent in an e-learning environment. Based on regulatory fit theory, this study also investigated if matching a learner’s regulatory focus orientation i.e., promotion-focus or prevention-focus to compatible agent’s message-frame i.e., gain-frame or loss-frame would produce cognitive and motivational benefits. The results of our experiment (n=210) revealed that the motivational virtual agent that utilized lossframe message to encourage learning behavior induced significantly higher germane cognitive load and intrinsic motivation in learners, as compared to the gain-frame motivational virtual agent. It was also shown that when the virtual agent used gain-frame message to encourage learning behavior, chronic promotion-focus learners experienced greater intrinsic motivation in e-learning than did chronic prevention-focus learners. Implications and suggestions for further research are discussed in this paper.
{"title":"Interacting With Motivational Virtual Agent: The Effects of Message Framing and Regulatory Fit in an E-Learning Environment","authors":"Tze Wei Liew, Su-Mae Tan, Chin Lay Gan","doi":"10.1109/ICDIM.2018.8846987","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8846987","url":null,"abstract":"Virtual agents can be integrated in e-learning environments to encourage learning behavior through persuasive and motivational messages. In this study, we aimed to investigate the effects of agent’s message-frame, i.e., gainframe and loss-frame on cognitive load and intrinsic motivation of learners interacting with motivational virtual agent in an e-learning environment. Based on regulatory fit theory, this study also investigated if matching a learner’s regulatory focus orientation i.e., promotion-focus or prevention-focus to compatible agent’s message-frame i.e., gain-frame or loss-frame would produce cognitive and motivational benefits. The results of our experiment (n=210) revealed that the motivational virtual agent that utilized lossframe message to encourage learning behavior induced significantly higher germane cognitive load and intrinsic motivation in learners, as compared to the gain-frame motivational virtual agent. It was also shown that when the virtual agent used gain-frame message to encourage learning behavior, chronic promotion-focus learners experienced greater intrinsic motivation in e-learning than did chronic prevention-focus learners. Implications and suggestions for further research are discussed in this paper.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128869568","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8846980
Markus Haberzettl, B. Markscheffel
The increase in daily emails sent to the customer service of companies is creating new challenges. Sentiment analysis, i.e. the automated recognition of mood and polarity in texts, is a solution to this problem, but the sentiment analysis of German emails is still an open research problem. With the help of a literature analysis we identify and analyze the most relevant machine learning methods and the corresponding feature extraction methods.
{"title":"A Literature Analysis for the Identification of Machine Learning and Feature Extraction Methods for Sentiment Analysis","authors":"Markus Haberzettl, B. Markscheffel","doi":"10.1109/ICDIM.2018.8846980","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8846980","url":null,"abstract":"The increase in daily emails sent to the customer service of companies is creating new challenges. Sentiment analysis, i.e. the automated recognition of mood and polarity in texts, is a solution to this problem, but the sentiment analysis of German emails is still an open research problem. With the help of a literature analysis we identify and analyze the most relevant machine learning methods and the corresponding feature extraction methods.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"166 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133209436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8847128
Burcu Kuleli Pak, Bora Mocan, Sema Yıldız Yoldaş, Neşe Baz
Digital marketing sector is an expanding sector with increasing number of customers, product and service types. Dynamic structure of websites, rapid changes in stock levels in ecommerce websites and dependency of existing systems to humans causes digital marketing agencies to require autonomous systems for customer account management. So the aim of this study is developing an autonomous intelligent system that operates integrated with Google Ads platform. The developed system makes optimization in return-on-investment, conversions, advertising texts and profit using Ackermann Feedback Control Algorithm and State Transition Matrices.
{"title":"Development of Autonomous Intelligent System for Google Ads","authors":"Burcu Kuleli Pak, Bora Mocan, Sema Yıldız Yoldaş, Neşe Baz","doi":"10.1109/ICDIM.2018.8847128","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8847128","url":null,"abstract":"Digital marketing sector is an expanding sector with increasing number of customers, product and service types. Dynamic structure of websites, rapid changes in stock levels in ecommerce websites and dependency of existing systems to humans causes digital marketing agencies to require autonomous systems for customer account management. So the aim of this study is developing an autonomous intelligent system that operates integrated with Google Ads platform. The developed system makes optimization in return-on-investment, conversions, advertising texts and profit using Ackermann Feedback Control Algorithm and State Transition Matrices.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132159331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8847136
G. Jakimovski, D. Davcev
Medical images (Magnetic Resonance Imaging scans) are used by doctors and medical specialists to determine the possibility that a cancer is present in the lungs of a patient. We are using these images, along with Deep Neural Network algorithms to help doctors with image diagnostics by training the Deep Neural Network (DNN) to recognize lung cancer. Our Deep Neural Network introduces novelty by making extensive search by adding additional layers of convolution and max pooling. Moreover, we are using images from slow progressing lung cancer to determine the threshold or at which point in the progression, our Deep Neural Network, will diagnose the cancer. Using this, doctors will have additional help in early phase lung cancer detection and early treatment. These are the main purposes of our research, which includes thorough search of possibilities of lung cancer and early detection.
{"title":"Lung cancer medical image recognition using Deep Neural Networks","authors":"G. Jakimovski, D. Davcev","doi":"10.1109/ICDIM.2018.8847136","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8847136","url":null,"abstract":"Medical images (Magnetic Resonance Imaging scans) are used by doctors and medical specialists to determine the possibility that a cancer is present in the lungs of a patient. We are using these images, along with Deep Neural Network algorithms to help doctors with image diagnostics by training the Deep Neural Network (DNN) to recognize lung cancer. Our Deep Neural Network introduces novelty by making extensive search by adding additional layers of convolution and max pooling. Moreover, we are using images from slow progressing lung cancer to determine the threshold or at which point in the progression, our Deep Neural Network, will diagnose the cancer. Using this, doctors will have additional help in early phase lung cancer detection and early treatment. These are the main purposes of our research, which includes thorough search of possibilities of lung cancer and early detection.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134512368","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8847057
H. Tellioglu
The digital transformation of our society is happening. In this paper, we try to provide means to deal with this phenomenon. We introduce and examine a new approach to show how to capture the impact of digital transformation methodically, and by doing so, how to guide the complex unpredictable process of digitalization in our social environment. After showing related work on artifacts, on the representation of things, on modeling, and finally on models as artifacts, we present our new model-based approach, the flow of models we developed, namely models for object characterization, hypothetical story, prediction, and test/experiment/evaluation. Furthermore, we show the context of our research, the role of models in design, and how we broaden our research context from design to digital transformation. Before we conclude our paper, we illustrate our approach on an example from health care, in the scope of an international research project.
{"title":"A Model-Based Approach to Guide Digital Transformation","authors":"H. Tellioglu","doi":"10.1109/ICDIM.2018.8847057","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8847057","url":null,"abstract":"The digital transformation of our society is happening. In this paper, we try to provide means to deal with this phenomenon. We introduce and examine a new approach to show how to capture the impact of digital transformation methodically, and by doing so, how to guide the complex unpredictable process of digitalization in our social environment. After showing related work on artifacts, on the representation of things, on modeling, and finally on models as artifacts, we present our new model-based approach, the flow of models we developed, namely models for object characterization, hypothetical story, prediction, and test/experiment/evaluation. Furthermore, we show the context of our research, the role of models in design, and how we broaden our research context from design to digital transformation. Before we conclude our paper, we illustrate our approach on an example from health care, in the scope of an international research project.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127109701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8846984
Lisa Ehrlinger, Thomas Grubinger, B. Varga, Mario Pichler, T. Natschläger, Jürgen Zeindl
With the advent of Industry 4.0, many companies aim at analyzing historically collected or operative transaction data. Despite the availability of large amounts of data, particular missing values can introduce bias or preclude the use of specific data analytics methods. Historically, a lot of research into missing data comes from the social sciences, especially with respect to survey data, whereas little research work deals with industrial missing data. In this paper, we (1) describe challenges that occur with missing data in the context of industrial data analytics, and (2) present an approach for handling missing data in industrial databases, which has been applied at voestalpine Stahl GmbH. In addition, we have evaluated different methods to impute missing values in our application data.
{"title":"Treating Missing Data in Industrial Data Analytics","authors":"Lisa Ehrlinger, Thomas Grubinger, B. Varga, Mario Pichler, T. Natschläger, Jürgen Zeindl","doi":"10.1109/ICDIM.2018.8846984","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8846984","url":null,"abstract":"With the advent of Industry 4.0, many companies aim at analyzing historically collected or operative transaction data. Despite the availability of large amounts of data, particular missing values can introduce bias or preclude the use of specific data analytics methods. Historically, a lot of research into missing data comes from the social sciences, especially with respect to survey data, whereas little research work deals with industrial missing data. In this paper, we (1) describe challenges that occur with missing data in the context of industrial data analytics, and (2) present an approach for handling missing data in industrial databases, which has been applied at voestalpine Stahl GmbH. In addition, we have evaluated different methods to impute missing values in our application data.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120909814","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The use of social media has grown as an essential communication tool for e-commerce. To be effective, a clear understanding of the impact of information on viewers should be reached. This paper investigates the impact of the different type of information on Facebook, i.e., detailed information, interactivity information and persuasive information, on the level of trust. Based on experiment design using an eye tracker devices, we find that detailed information and interactivity information have a positive and significant effect on trust. Therefore, information about the product and promise for quick response to the question are critical to be included on a Facebook page.
{"title":"The Effect of Different Type of Information on Trust in Facebook Page","authors":"Hasrini Sari, Farhan Mutaqin, Aditya Parama Setiaboedi","doi":"10.1109/ICDIM.2018.8846978","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8846978","url":null,"abstract":"The use of social media has grown as an essential communication tool for e-commerce. To be effective, a clear understanding of the impact of information on viewers should be reached. This paper investigates the impact of the different type of information on Facebook, i.e., detailed information, interactivity information and persuasive information, on the level of trust. Based on experiment design using an eye tracker devices, we find that detailed information and interactivity information have a positive and significant effect on trust. Therefore, information about the product and promise for quick response to the question are critical to be included on a Facebook page.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114454750","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8847139
Christian Grévisse, J. Meder, J. Botev, S. Rothkugel
Document collections in e-learning can cause issues to both learners and teachers. On one hand, inquiry from the vast corpus of available resources is non-trivial without adequate formulation support and semantic information. Implicit links between documents are hardly understood without a proper visualization. On the other hand, it is difficult for teachers to keep track of the topics covered by a large collection. In this paper, we present an ontology coverage tool and document browser for learning material exploration. Both learners and teachers can benefit from a visualization of an ontology and the documents related to the comprised concepts, overcoming the limitations of traditional file explorers. Guiding users through a visual query process, learners can quickly pinpoint relevant learning material. The visualization, which has been implemented as a web application using the D3.js JavaScript library, can be integrated into different e-learning applications to further enhance the workflow of learners. Finally, teachers are provided an overview of topic coverage within the collection.
{"title":"Ontology Coverage Tool and Document Browser for Learning Material Exploration","authors":"Christian Grévisse, J. Meder, J. Botev, S. Rothkugel","doi":"10.1109/ICDIM.2018.8847139","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8847139","url":null,"abstract":"Document collections in e-learning can cause issues to both learners and teachers. On one hand, inquiry from the vast corpus of available resources is non-trivial without adequate formulation support and semantic information. Implicit links between documents are hardly understood without a proper visualization. On the other hand, it is difficult for teachers to keep track of the topics covered by a large collection. In this paper, we present an ontology coverage tool and document browser for learning material exploration. Both learners and teachers can benefit from a visualization of an ontology and the documents related to the comprised concepts, overcoming the limitations of traditional file explorers. Guiding users through a visual query process, learners can quickly pinpoint relevant learning material. The visualization, which has been implemented as a web application using the D3.js JavaScript library, can be integrated into different e-learning applications to further enhance the workflow of learners. Finally, teachers are provided an overview of topic coverage within the collection.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114473348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-09-01DOI: 10.1109/ICDIM.2018.8847003
B. Murray, L. Perera
Autonomous vehicles will be an integral part of future transportation systems, and the maritime industry is working towards developing methods to ensure safe autonomous ship operations. One of the major challenges in realizing autonomous ships is ensuring effective collision avoidance technologies. Autonomous vessels must have a higher degree of situation awareness to detect other vessels, predict their future intentions, and evaluate the respective collision risk. One step in achieving this goal is to predict other vessel trajectories accurately. In this paper, a data-driven approach to vessel trajectory prediction for time horizons of 5–30 minutes utilizing historical AIS data is evaluated. A clustering based Single Point Neighbor Search Method is investigated along with a novel Multiple Trajectory Extraction Method. Predictions have been conducted using these methods and compared with the Constant Velocity Method. Additionally, the Multiple Trajectory Extraction Method is utilized to evaluate estimated ship routes.
{"title":"A Data-Driven Approach to Vessel Trajectory Prediction for Safe Autonomous Ship Operations","authors":"B. Murray, L. Perera","doi":"10.1109/ICDIM.2018.8847003","DOIUrl":"https://doi.org/10.1109/ICDIM.2018.8847003","url":null,"abstract":"Autonomous vehicles will be an integral part of future transportation systems, and the maritime industry is working towards developing methods to ensure safe autonomous ship operations. One of the major challenges in realizing autonomous ships is ensuring effective collision avoidance technologies. Autonomous vessels must have a higher degree of situation awareness to detect other vessels, predict their future intentions, and evaluate the respective collision risk. One step in achieving this goal is to predict other vessel trajectories accurately. In this paper, a data-driven approach to vessel trajectory prediction for time horizons of 5–30 minutes utilizing historical AIS data is evaluated. A clustering based Single Point Neighbor Search Method is investigated along with a novel Multiple Trajectory Extraction Method. Predictions have been conducted using these methods and compared with the Constant Velocity Method. Additionally, the Multiple Trajectory Extraction Method is utilized to evaluate estimated ship routes.","PeriodicalId":120884,"journal":{"name":"2018 Thirteenth International Conference on Digital Information Management (ICDIM)","volume":"162 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121558902","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：