Romanian Journal of
Human - Computer InteractionVol.9, No.3, 2016
ISSN 1843-4460
Contents
Methods for Modelling Sketches in the Collaborative Prototyping of User Interfaces
Jorge Luis Pérez Medina
183 - 216 Recovering implicit thread structure in chat conversations
Andrei Dulceanu
217 - 232 Aggregating textual and video data from movies
Alexandru Hulea, Traian Rebedea
233 - 254 Recommendation Technique for the “Cold-Start” Problem
Mihaela Colhon, Adrian Iftene
255 - 268
Abstracts
Methods for Modelling Sketches in the Collaborative Prototyping of User Interfaces
Jorge Luis Pérez Medina
Université catholique de Louvain, Louvain School of Management Research Institute
Place des Doyens, 1 – B-1348 Louvain-la-Neuve (Belgium)
E-mail:jorge.perezmedina@uclouvain.be
Abstract: Cross-functional teams with different technical backgrounds working on cross-platform environments require the production of flexible modeling of user interfaces in early steps of a design process. We observe that model-driven engineering (MDE) is currently gaining acceptance in many domains. However, existing solutions have no support for collaborative prototyping of user interfaces by sketching recognition for multiple stakeholders (e.g., designers, developers, final end users) working with heterogeneous computing platforms (e.g., smartphones, tablets, laptops, desktop), on different, perhaps separate or shared, interaction surfaces (e.g., tables, whiteboards) in a co-located way or remotely. This requires flexibility to explore and reuse vague and uncertain ideas as model sketching. This paper presents UsiSketch, an MDE method for modeling sketches that offers the following novel features resulting from a requirement elicitation process: sketching recognition on different surfaces based on a new recognition algorithm that accommodates very large surfaces and model-based design of user interfaces with collaboration.
Keywords: Sketching, Collaborative Prototyping, Graphical User Interface, Design Tools and Techniques.
Cite this paper as:
Pérez Medina, J. L. Methods for Modelling Sketches in the Collaborative Prototyping of User Interfaces. Revista Romana de Interactiune Om-Calculator 9(3), 183-216, 2016.
Recovering implicit thread structure in chat conversations
Andrei Dulceanu
Politehnica University of Bucharest, Romania
313 Splaiul Independenței, 060042, Bucharest, Romania
E-mail: andrei.dulceanu@gmail.com
Abstract: The analysis of chat conversations is a cumbersome task because of the number of different discussion threads that may occur at a certain moment. While most participants in a chat session tend to discuss one topic at a time, interferences appear due to environment asynchrony. This paper presents an approach for recovering implicit thread structure of a chat conversation by using a pipeline centered on semantic similarity between short phrases. Temporal, social and lexical aspects of the conversation are blended in a single model which predicts for each utterance not only the thread it belongs to, but also the utterance most related to in its thread.
Keywords: chat conversation, speech act, thread, disentanglement, semantic similarity, WordNet.
Cite this paper as:
Dulceanu, A. Recovering implicit thread structure in chat conversations. Revista Romana de Interactiune Om-Calculator 9(3), 217-232, 2016.
Aggregating textual and video data from movies
Alexandru Hulea, Traian Rebedea
University Politehnica of Bucharest
Splaiul Independenței nr. 313, sector 6, 060042, Bucuresti
E-mail: alexandru.hulea@gmail.com, traian.rebedea@cs.pub.roAbstract: In this paper, we present an automatically annotated corpus based on movie screenplays (script) and subtitles. We extract the relevant textual information from movie screenplays and subtitles using a regular expression approach. Then, we synchronize screenplays with subtitles using a matching algorithm, thus bounding each sentence from a script between two temporal limits. We also developed an application using the corpus to test our approach and to show practical situations where this corpus is useful. The application employs topic detection and it involves searching for a specified topic in the movie text and marking the topic as non-existent, episodic or primary topic for the analyzed text. The major problem we faced while working on this system was the unexpected structure of the screenplay sheets as this kind of files are not entirely written using a standardized format which can be easily parsed and structured automatically. Some types of errors can be overcome with regular expressions, but there are other errors that need a machine learning approach to be surpassed.
Keywords: Video-text annotation, Corpus creation, Video understanding, Topic detection, Information retrieval.
Cite this paper as:
Hulea, A., Rebedea, T. Aggregating textual and video data from movies. Revista Romana de Interactiune Om-Calculator 9(3), 233-254, 2016.
Recommendation Technique for the “Cold-Start” Problem
Mihaela Colhon1, Adrian Iftene2
1University of Craiova, Department of Computer Science
Street Alexandru Ioan Cuza, Nr. 13, 200585, Craiova
E-mail: mcolhon@inf.ucv.ro
2 Alexandru Ioan Cuza University of Iaşi, Faculty of Computer Science Blvd. Carol I, no. 11, Iaşi
E-mail: adiftene@inf.uaic.roAbstract: Usually, users that benefit from a Recommender System outputs only get a list of items that the system assumes best match their needs, without having any clue regarding how the system managed to figure out what they like. In this paper, we propose a mechanism that generates content-based recommendations organized on levels of similarities with the selected product in order to let the user decide about what similarity degree (s)he wants to explore. We demonstrate empirically that our proposed mechanism can ensure good performance for a recommendation technique under the cold-start conditions.
Keywords: human-computer interaction, recommender system, content-based filtering, rough set theory.
Cite this paper as:
Colhon, M., Iftene, A. Recommendation Technique for the “Cold-Start” Problem. Revista Romana de Interactiune Om-Calculator 9(3), 255-268, 2016.