1,720,960 research outputs found
Online resource for the article Artificial Neural Network based audio reinforcement for computer assisted rote learning
No full textOnline resource for the article Artificial Neural Network based audio reinforcement for computer assisted rote learning</p
Displaying visual stimuli and recording audio
No full textThis repository contains the custom codes and resources
for building the audio stimuli used in the experiments reported in our
article. The implementation is done in Python 3.5.2 and it deploys
several functions from the package pydub 0.25.1.
We created two main routines: (i) for displaying the image
stimuli to the "speaker" and recording his voice and (ii) for
post-processing the raw recordings to detect silent segments, unify them
and create audio clips to be used in e-learning experiments.
Displaying image stimuli and recording speaker's voice
The program show_and_record_cound.py displays visual
stimuli (images) involving text or numbers to a person (i.e. "speaker")
and records his/her voice into an audio clip, while he/she reads aloud
the visual stimuli.
The speaker is notified with a brief beep sound at the
beginning of each image, which is displayed for a time window of 4
seconds. After the notification, audio recordings starts and goes on
until the end of the time window. As the time window is finished, an
audio clip is saved. Subsequently, the current image is flushed and the
next image (e.g. a word or number) is displayed. The outputs of the
program are (i) an audio clip for each image and (ii) the
list of words/numbers appearing on the images stored in a text file.
Post-processing of raw audio recordings
The program silence_detect.py receives a list of raw audio
recordings and distinguishes the voice and silent segments in each of
them. It then builds an audio clip by appending a recording (i.e.
country_name) with another one (i.e. capital_name) by padding a brief
artificial silent segment in between. It repeats this sequence
once more, appends to other silent segments at the beginning and at the
end, and finally stores the result into an audio clip.</p
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Supplemental material for Artificial Neural Network based audio reinforcement for computer assisted rote learning
No full textSupplemental material for Artificial Neural Network based audio reinforcement for computer assisted rote learning
ANN-based audio reinforcement for computer assisted rote learning
No full textThis repository contains the custom codes and tools developed to obtain the results reported in our article [1]:
Note that the raw data set contains privacy sensitive information, which is concealed in this release.
Activity log files
The activity log files are recorded from the e-learning
software Anki [2]. They contain three basic kinds of information as
temporal, identifier and evaluation.
The temporal variables are registered in UNIX time at
millisecond resolution and include tp, tf, and te. Here, tp denotes the
time of prompt, i.e. the instant when the Q-face of a card appears. In
addition, tf represents the time of flip, i.e. the instant when the
learner presses the ''Show Answer'' button and discloses the A-face of
the card. Finally, te stands for the time of evaluation. Namely, it is
the instant when the learner assesses the difficulty of a card by
choosing one of ''Again'', ''Good'' or ''Easy'', which is also
registered as the evaluation variable. For more detail, please see https://docs.ankiweb.net/#/ .
On the other hand, the identifier variables are integer
codes used to determine the deck or card that is being studied
(i.e.displayed) at a given time instant (e.g. deck ID, card ID). Note
that each log file is associated with a single user. Namely, the
software recorded one log file into the account of each user. In
addition, each line of the activity log file corresponds to a single
action of the user which is considered as a reaction to the software.
The structure of each line of data is as follows: [UNIX time], function
name, data in detail (i.e. flags, queue).
Memory score file
This file stores the performance score of the participants
in the prior (P), short-term (S) and mid-term (M) memory tests. The
file contains 6 sheets, 3 for the exploration stage and 3 for the
verification stage. Concerning each stage, the scores concerning P, S
and M tests are stored in separate sheets.
Regarding the coding of memory test scores, if the
participant recalls the answer of a certain query successfully, we
register the score of that card with 1, otherwise 0. Different columns
contain the scores of different participants and different rows contain
the information concerning different cards. In addition, the card ID,
Q-face and A-face information is also written on the left hand side of
each row as a key for matching with the activity log files.
Model development
The function NN_model_builder.py is used for building the
estimator. We considered building a model for every learning task.
First, the input data file (compressed as pickle files) are loaded and
the different behavioral variables stored in each column are separated.
We drop the column, which is not used for building the model. Next,
StandardScaler is applied to the retained data. Subsequently, we reshape
the train set and convert the labels to categorical data before feeding
them into the model. Finally, we develop an artificial neural network
which serves as a model for estimation of memory performance as
improvement or deterioration. We test the accuracy of the model and save
them into the pickle files ready to be imported by the e-learning
software.
Integration of the model to the e-learning software
The function NN_model_integration.py is integrated with
the e-learning software for estimating memory performance in an
on-the-fly manner. First, the behavioral variables are calculated.
Should the operating mode of the e-learning software be
'Estimation', the variables are then prepared in the same way as
described above. The pre-trained estimation model is loaded and the
prepared variables are fed into it as inputs. Depending on the
estimation result, the audio key (i.e. trigger or no-trigger) is set and
stored in a log file.
Should the operating mode of the e-learning software be
'Full audio reinforcement',the audio key is set to 'ON' (constant
triggering of audio). The audio trigger log file is updated accordingly.
References
[1] ANN-based audio reinforcement for computer assisted rote learning,
Authors, Under review
[2] D. Elmes, “Anki - friendly, intelligent flashcards.” https://ankiweb.net/about, 2021.</p
Variations on the Author
Get PDF“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Get PDFWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Dispelling the Myths Behind First-author Citation Counts
Get PDFWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
- …
