-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathfeed.xml
363 lines (304 loc) · 25.6 KB
/
feed.xml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom" ><generator uri="https://jekyllrb.com/" version="4.0.1">Jekyll</generator><link href="https://bvasiles.github.io/feed.xml" rel="self" type="application/atom+xml" /><link href="https://bvasiles.github.io/" rel="alternate" type="text/html" /><updated>2023-02-14T03:40:40+00:00</updated><id>https://bvasiles.github.io/feed.xml</id><title type="html">Bogdan Vasilescu’s homepage</title><subtitle>Bogdan Vasilescu's homepage at Carnegie Mellon University</subtitle><author><name>Bogdan Vasilescu</name></author><entry><title type="html">Conference Monitor</title><link href="https://bvasiles.github.io/conference-monitor/" rel="alternate" type="text/html" title="Conference Monitor" /><published>2015-04-05T00:00:00+00:00</published><updated>2015-04-05T00:00:00+00:00</updated><id>https://bvasiles.github.io/conference-monitor</id><content type="html" xml:base="https://bvasiles.github.io/conference-monitor/"><p>Recently I started experimenting with <a href="http://www.highcharts.com">Highcharts</a>
to visualize <a href="http://bvasiles.github.io/papers/scico13.pdf">our conference metrics</a>
in the browser. Here’s a sample. Conference organizers may find these useful.
Click (tap) on conference names in the legend below each plot to enable those
series. The data is stored in <a href="https://github.com/tue-mdse/conferenceMetrics">this GitHub repo</a>
and loaded dynamically on page load.</p>
<h3 id="acceptance-rates">Acceptance rates</h3>
<h5 id="submissions">Submissions</h5>
<p>Number of papers submitted to the <em>main research track</em>.
<strong>Update:</strong> For MSR this is the number of <em>full</em> papers submitted to the
main research track. Thanks to <a href="http://thomas-zimmermann.com">Tom Zimmermann</a>
for the clarification.</p>
<div id="container_sp" style="min-width:310px; height:400px; margin:0 auto; width:800px;"></div>
<!-- (without distinguishing between long and short papers, if both were part of the main track). -->
<h5 id="acceptance-rates-1">Acceptance rates</h5>
<p>Ratio between the number of accepted full papers and the number of submissions.</p>
<div id="container_ra" style="min-width:310px; height:400px; margin:0 auto; width:800px;"></div>
<h3 id="program-committees">Program committees</h3>
<h5 id="pc-size">PC size</h5>
<p>Number of PC members.</p>
<div id="container_c" style="min-width:310px; height:400px; margin:0 auto; width:800px;"></div>
<h5 id="review-load">Review load</h5>
<p>Ratio between number of submissions and number of PC members.
If each submission is reviewed by 3 PC members, multiply each value by 3 to
see how many submissions are assigned to each PC member.</p>
<div id="container_rl" style="min-width:310px; height:400px; margin:0 auto; width:800px;"></div>
<h3 id="pc-turnover">PC turnover</h3>
<p>Fraction of PC members in a given year which are new with respect to the
previous year.</p>
<div id="container_rnc1" style="min-width:310px; height:400px; margin:0 auto; width:800px;"></div>
<h3 id="openness">Openness</h3>
<h5 id="papers-by-new-authors-wrt-the-previous-4-editions">Papers by new authors (w.r.t. the previous 4 editions)</h5>
<p>Fraction of accepted papers that have been co-authored by <em>new authors</em>, i.e.,
papers for which none of the authors published at any of the previous 4 editions
of this conference.</p>
<div id="container_rpna4" style="min-width:310px; height:400px; margin:0 auto; width:800px;"></div>
<h5 id="papers-by-pc-members">Papers by PC members</h5>
<p>Fraction of accepted papers that have been co-authored by at least one PC member.</p>
<div id="container_rac0" style="min-width:310px; height:400px; margin:0 auto; width:800px;"></div>
<h4 id="note">Note</h4>
<p>You can contribute additional data by submitting pull requests to
<a href="https://github.com/tue-mdse/conferenceMetrics">this repository</a>.</p>
<p>More details about the metrics:</p>
<blockquote>
<p><a href="http://bvasiles.github.io/papers/scico13.pdf">How healthy are software engineering conferences?</a>
<em>Vasilescu, B., Serebrenik, A., Mens, T., Brand, M.G.J. van den, and Pek, E.</em>
Science of Computer Programming 89, Part C, (2014), 251–272.</p>
</blockquote>
<script src="http://code.highcharts.com/highcharts.js"></script>
<script src="http://code.highcharts.com/modules/data.js"></script>
<script src="http://code.highcharts.com/modules/exporting.js"></script>
<script type="text/javascript" src="http://ajax.googleapis.com/ajax/libs/jquery/1.8.2/jquery.min.js"></script>
<!-- Additional files for the Highslide popup effect -->
<script type="text/javascript" src="http://www.highcharts.com/media/com_demo/highslide-full.min.js"></script>
<script type="text/javascript" src="http://www.highcharts.com/media/com_demo/highslide.config.js" charset="utf-8"></script>
<link rel="stylesheet" type="text/css" href="http://www.highcharts.com/media/com_demo/highslide.css" />
<script type="text/javascript">
$(function () {
function myChart(myTitle, myYLbl, csvPath, container) {
var options = {
chart: {
renderTo: container,
type: 'line'
},
title: {
text: myTitle
},
xAxis: {
categories: []
},
yAxis: {
title: {
text: myYLbl
}
},
series: [],
tooltip: {
formatter: function() {
var s = [];
$.each(this.points, function(i, point) {
s.push('<span class="tooltip">' + point.series.name + ' : ' +
point.y + '<br><span>');
});
return s.join('');
},
shared: true
}
};
// Get the CSV and create the chart
$.get(csvPath, function(csv) {
var dataColumns = [];
var dataNames = [];
var xLabels = [];
// Split the lines
var lines = csv.split('\n');
// Extract categories from header
// year,ICSE,ICSM,WCRE,CSMR,MSR,GPCE,FASE,ICPC,FSE,SCAM,ASE
var header = lines[0];
var items = header.split(';');
for (i=1;i<items.length;i++) {
dataColumns.push([]);
}
dataNames = items.slice(1,items.length);
// Extract data for the most recent 10 years
// 2013,50,71,65,64,46,27,30,50,37,36,33
// 2012,48,70,65,64,41,32,31,38,36,42,44
var rlines = lines.slice(1, 11);
rlines.reverse();
$.each(rlines, function (lineNo, line) {
var items = line.split(';');
$.each(items, function (itemNo, item) {
// Data is from column 1 onwards
if (itemNo > 0) {
if (item.length == 0){
var val = null;
}else{
var val = parseFloat(item);
}
dataColumns[itemNo-1].push(val);
}
// The year is column 0
else{
xLabels.push(item);
options.xAxis.categories.push(item);
}
});
});
// Create data series
for (i=0;i<dataColumns.length;i++) {
var r = dataColumns[i];
var series = {
data: r,
name: dataNames[i]
};
if (series.name !== "ICSE" &
series.name !== "FSE" &
series.name !== "ICSM") {
series.visible = false;
}
options.series.push(series);
}
var chart = new Highcharts.Chart(options);
});
}
myChart('Submissions',
'#Submitted_papers',
'https://gist.githubusercontent.com/bvasiles/25b581b408928c547e21/raw/SP.csv',
'container_sp');
myChart('Acceptance Rates',
'#Accepted_papers / #Submitted_papers',
'https://gist.githubusercontent.com/bvasiles/25b581b408928c547e21/raw/RA.csv',
'container_ra');
myChart('Program Committee Size',
'#PC members',
'https://gist.githubusercontent.com/bvasiles/25b581b408928c547e21/raw/C.csv',
'container_c');
myChart('PC Turnover',
'Fraction PC new w.r.t. previous year',
'https://gist.githubusercontent.com/bvasiles/25b581b408928c547e21/raw/RNC1.csv',
'container_rnc1');
myChart('Review Load (x 3)',
'#Submissions / #PC_members',
'https://gist.githubusercontent.com/bvasiles/25b581b408928c547e21/raw/RL.csv',
'container_rl');
myChart('Papers by New Authors (w.r.t. previous 4 years)',
'Fraction papers by new authors',
'https://gist.githubusercontent.com/bvasiles/25b581b408928c547e21/raw/RPNA4.csv',
'container_rpna4');
myChart('Fraction Papers by PC members',
'Fraction PC papers',
'https://gist.githubusercontent.com/bvasiles/25b581b408928c547e21/raw/RAC0.csv',
'container_rac0');
});
</script></content><author><name>Bogdan Vasilescu</name></author><summary type="html">Recently I started experimenting with Highcharts to visualize our conference metrics in the browser. Here’s a sample. Conference organizers may find these useful. Click (tap) on conference names in the legend below each plot to enable those series. The data is stored in this GitHub repo and loaded dynamically on page load. Acceptance rates Submissions Number of papers submitted to the main research track. Update: For MSR this is the number of full papers submitted to the main research track. Thanks to Tom Zimmermann for the clarification. Acceptance rates Ratio between the number of accepted full papers and the number of submissions. Program committees PC size Number of PC members. Review load Ratio between number of submissions and number of PC members. If each submission is reviewed by 3 PC members, multiply each value by 3 to see how many submissions are assigned to each PC member. PC turnover Fraction of PC members in a given year which are new with respect to the previous year. Openness Papers by new authors (w.r.t. the previous 4 editions) Fraction of accepted papers that have been co-authored by new authors, i.e., papers for which none of the authors published at any of the previous 4 editions of this conference. Papers by PC members Fraction of accepted papers that have been co-authored by at least one PC member. Note You can contribute additional data by submitting pull requests to this repository. More details about the metrics: How healthy are software engineering conferences? Vasilescu, B., Serebrenik, A., Mens, T., Brand, M.G.J. van den, and Pek, E. Science of Computer Programming 89, Part C, (2014), 251–272.</summary></entry><entry><title type="html">CHI’15 Press Release</title><link href="https://bvasiles.github.io/chi-press/" rel="alternate" type="text/html" title="CHI'15 Press Release" /><published>2015-01-19T00:00:00+00:00</published><updated>2015-01-19T00:00:00+00:00</updated><id>https://bvasiles.github.io/chi-press</id><content type="html" xml:base="https://bvasiles.github.io/chi-press/"><p><strong>Having higher gender and tenure diversity is associated with higher
productivity in <a href="http://github.com">GitHub</a> teams</strong>.</p>
<p>A joint team of computer science researchers from
<a href="http://www.cs.ucdavis.edu">University of California, Davis</a> and
<a href="http://www.tue.nl/en/">Eindhoven University of Technology</a>, in The Netherlands,
have conducted a study of the effects of gender and tenure diversity on
productivity and turnover in teams of Open Source Software (OSS) developers,
and have found small but positive effects.
The paper describing the results has been accepted for presentation at the
prestigious <a href="http://chi2015.acm.org">2015 ACM CHI Conference on Human Factors
in Computing Systems</a>, in Seoul, South Korea, in April 2015.
The paper was authored by
<a href="http://bvasiles.github.io">Bogdan Vasilescu</a>,
<a href="http://scholar.google.com/citations?user=IT0VNZkAAAAJ&amp;hl=en">Daryl Posnett</a>,
<a href="http://baishakhir.github.io">Baishakhi Ray</a>,
<a href="http://www.win.tue.nl/~mvdbrand/">Mark van den Brand</a>,
<a href="http://www.win.tue.nl/~aserebre">Alexander Serebrenik</a>,
<a href="http://www.cs.ucdavis.edu/~devanbu/">Prem Devanbu</a>,
and <a href="http://www.cs.ucdavis.edu/~filkov/">Vladimir Filkov</a>.</p>
<p>Dr. Vasilescu, the first author on the paper, and colleagues used regression
modeling on carefully-extracted data from more than 23,000 projects on GitHub,
the largest and most popular online collaborative coding platform.
To triangulate the findings, they also ran a user survey, with more than 800
respondents.
Their models show that after controlling for team size and other confounds
(such as a project’s age, development model, or amount of social activity),
both <em>gender and tenure diversity are positive and significant predictors of
productivity</em>, together explaining a small but significant fraction of the data
variability.
The benefits of experience diversity have a limit, though, as higher tenure
diversity may increase attrition.
This negative effect appears to be mitigated, however, when more experienced
people are present.
Survey respondents acknowledged the importance of diversity on their team’s
functioning, but reported different perceived effects.</p>
<p>This is the first academic study to consider effects of <em>gender</em> diversity on
productivity and turnover in OSS communities.
Based on these findings, Prof. Filkov, the senior author on the paper, believes
that increasing gender and tenure diversity in software teams may benefit
productivity, if all other confounds are equal.
Traditionally, women are significantly underrepresented in OSS and other
technical teams.
On a larger, economic and societal scale, these findings suggest that added
investments in educational and professional training efforts and outreach for
female programmers will likely result in added overall value.</p>
<p>A preprint containing more details is available
<a href="http://bvasiles.github.io/papers/chi15.pdf">here</a>.
Dr. Vasilescu can be reached at [email protected].
Prof. Filkov can be reached at [email protected].</p></content><author><name>Bogdan Vasilescu</name></author><summary type="html">Having higher gender and tenure diversity is associated with higher productivity in GitHub teams.</summary></entry><entry><title type="html">Gender and tenure diversity in GitHub teams</title><link href="https://bvasiles.github.io/gender-tenure-diversity-github/" rel="alternate" type="text/html" title="Gender and tenure diversity in GitHub teams" /><published>2015-01-19T00:00:00+00:00</published><updated>2015-01-19T00:00:00+00:00</updated><id>https://bvasiles.github.io/gender-tenure-diversity-github</id><content type="html" xml:base="https://bvasiles.github.io/gender-tenure-diversity-github/"><p>Diversity in teams arises from any attribute that people use to differentiate
themselves from others.
The most obvious attributes are arguably demographic (age, gender, culture,
ethnicity), but they can also be related to pretty much everything else (for
example, role, tenure, expertise, or even personality).
In general, diversity in teams is viewed as a double-edged sword.
On the one hand, increased team diversity results in more varied backgrounds
and ideas, which provide the team with access to broader information, enhanced
creativity, adaptability, and problem solving skills.
On the other hand, due to greater perceived differences in values, norms, and
communication styles in more diverse teams, members become more likely to
engage in stereotyping, cliquishness, and conflict.</p>
<p>Diversity in teams has been studied for a long time in offline groups, but
different studies still disagree on the effects.
Instead, we focused on distributed (online) software teams, such as those in
Open Source Software (OSS).
OSS teams are much more fluid, therefore much less tangible, than their
offline counterparts.
In OSS teams are <em>naturally</em> very diverse, consisting of contributors from all
over the world, typically a mixture of volunteers and professionals, coming
from varied cultural and educational backgrounds, with different
interests and skills.</p>
<p>We focused on two diversity attributes that are prominent in OSS: <strong>gender</strong>
and <strong>tenure</strong> (experience).
Women are underrepresented in programming, and especially so in OSS.
Moreover, the “hacker” culture is said to be male-dominated and unfriendly
to women, with reports of active discrimination and sexism.
OSS teams are inherently diverse with respect to experience, since they often
rely on a steady influx of new contributors.</p>
<p>Then, we carefully extracted data from more than 23,000 active collaborative
projects on <a href="http://github.com">GitHub</a>, the largest and most popular online
collaborative coding platform.
Each observation in this data set contains the composition, characteristics,
and outcomes of a project’s team of contributors for each quarter (90-day
period) in the evolution of the project.
Using regression analysis on this data, we modeled:</p>
<ul>
<li><em>productivity</em> (the number of commits by team developers recorded in either
the main repository or any of its forks in a given quarter), and</li>
<li><em>turnover</em> (the fraction of the team in a given quarter that is different
with respect to previous quarter)</li>
</ul>
<p>as functions of <em>gender diversity</em> (measured using the Blau diversity index)
and <em>tenure diversity</em> (measured using the coefficient of variation).
We controlled for many confounds:</p>
<ul>
<li><em>team size</em> (larger teams are likely associated with increased productivity
and increased turnover);</li>
<li><em>overall project activity</em> (total number of commits);</li>
<li><em>project forks</em> (activity in forks is more likely to be limited, both in
time and in amount, compared to a project’s main repository);</li>
<li><em>time</em> (has a moderating influence on the effects of
diversity: as
group members continue to interact, they have more time to
adjust to differences between them);</li>
<li><em>project age</em> (projects that started later and their teams may have
experienced a different GitHub culture);</li>
<li><em>tenure median</em> (to distinguish between more and less experienced teams,
regardless of how diverse these teams are with respect to tenure);</li>
<li><em>comments</em> (number of GitHub comments during a given quarter on commits,
pull requests, and issues, as a reflection of a project’s social activity).</li>
</ul>
<p>Our models show that both <strong>gender and tenure diversity are positive and
significant predictors of productivity</strong>, together explaining a small but
significant fraction of the data variability.</p>
<p>The paper presenting these results (co-authored by
<a href="http://bvasiles.github.io">Bogdan Vasilescu</a>,
<a href="http://scholar.google.com/citations?user=IT0VNZkAAAAJ&amp;hl=en">Daryl Posnett</a>,
<a href="http://baishakhir.github.io">Baishakhi Ray</a>,
<a href="http://www.win.tue.nl/~mvdbrand/">Mark van den Brand</a>,
<a href="http://www.win.tue.nl/~aserebre">Alexander Serebrenik</a>,
<a href="http://www.cs.ucdavis.edu/~devanbu/">Prem Devanbu</a>,
and <a href="http://www.cs.ucdavis.edu/~filkov/">Vladimir Filkov</a>) has been accepted
for presentation at the <a href="http://chi2015.acm.org">2015 ACM CHI Conference on Human Factors
in Computing Systems</a>, in Seoul, South Korea, in
April 2015.
A preprint containing more details is available
<a href="/papers/chi15.pdf">here</a>.</p>
<p>This is the first academic study to consider effects of <strong>gender</strong> diversity on
productivity and turnover in OSS communities.
On a larger, economic and societal scale, these findings suggest that added
investments in educational and professional training efforts and outreach for
female programmers will likely result in added overall value.</p></content><author><name>Bogdan Vasilescu</name></author><summary type="html">Diversity in teams arises from any attribute that people use to differentiate themselves from others. The most obvious attributes are arguably demographic (age, gender, culture, ethnicity), but they can also be related to pretty much everything else (for example, role, tenure, expertise, or even personality). In general, diversity in teams is viewed as a double-edged sword. On the one hand, increased team diversity results in more varied backgrounds and ideas, which provide the team with access to broader information, enhanced creativity, adaptability, and problem solving skills. On the other hand, due to greater perceived differences in values, norms, and communication styles in more diverse teams, members become more likely to engage in stereotyping, cliquishness, and conflict. Diversity in teams has been studied for a long time in offline groups, but different studies still disagree on the effects. Instead, we focused on distributed (online) software teams, such as those in Open Source Software (OSS). OSS teams are much more fluid, therefore much less tangible, than their offline counterparts. In OSS teams are naturally very diverse, consisting of contributors from all over the world, typically a mixture of volunteers and professionals, coming from varied cultural and educational backgrounds, with different interests and skills. We focused on two diversity attributes that are prominent in OSS: gender and tenure (experience). Women are underrepresented in programming, and especially so in OSS. Moreover, the “hacker” culture is said to be male-dominated and unfriendly to women, with reports of active discrimination and sexism. OSS teams are inherently diverse with respect to experience, since they often rely on a steady influx of new contributors. Then, we carefully extracted data from more than 23,000 active collaborative projects on GitHub, the largest and most popular online collaborative coding platform. Each observation in this data set contains the composition, characteristics, and outcomes of a project’s team of contributors for each quarter (90-day period) in the evolution of the project. Using regression analysis on this data, we modeled: productivity (the number of commits by team developers recorded in either the main repository or any of its forks in a given quarter), and turnover (the fraction of the team in a given quarter that is different with respect to previous quarter) as functions of gender diversity (measured using the Blau diversity index) and tenure diversity (measured using the coefficient of variation). We controlled for many confounds: team size (larger teams are likely associated with increased productivity and increased turnover); overall project activity (total number of commits); project forks (activity in forks is more likely to be limited, both in time and in amount, compared to a project’s main repository); time (has a moderating influence on the effects of diversity: as group members continue to interact, they have more time to adjust to differences between them); project age (projects that started later and their teams may have experienced a different GitHub culture); tenure median (to distinguish between more and less experienced teams, regardless of how diverse these teams are with respect to tenure); comments (number of GitHub comments during a given quarter on commits, pull requests, and issues, as a reflection of a project’s social activity). Our models show that both gender and tenure diversity are positive and significant predictors of productivity, together explaining a small but significant fraction of the data variability. The paper presenting these results (co-authored by Bogdan Vasilescu, Daryl Posnett, Baishakhi Ray, Mark van den Brand, Alexander Serebrenik, Prem Devanbu, and Vladimir Filkov) has been accepted for presentation at the 2015 ACM CHI Conference on Human Factors in Computing Systems, in Seoul, South Korea, in April 2015. A preprint containing more details is available here. This is the first academic study to consider effects of gender diversity on productivity and turnover in OSS communities. On a larger, economic and societal scale, these findings suggest that added investments in educational and professional training efforts and outreach for female programmers will likely result in added overall value.</summary></entry></feed>