Grants:APG/Proposals/2020-2021 round 2/Wikimedia Indonesia/Progress report form

Introduction

Wikimedia Indonesia is willing to cover more than 700 languages and 300 ethnic groups across the island nations of Indonesia to get them available on the internet. Our five-year strategic planning from 2017-2022 emphasizes on community development and support to achieve our goals. We believe by having a stronger community presence, we can build a more sustainable environment for nurturing the free knowledge culture. Although we are still in the pandemic situation, we try to be creative as possible to build new communities, support, and strengthen the strengthen the established communities. We also invite and reach various organisations to get involved in Wikimedia movements and raise awareness about the importance of free knowledge. We believe by reaching out to various communities and organisations, Indonesian local content will be more enriched and our movement can be more inclusive. These efforts are in line with knowledge equity and to make the knowledge that we provided as a primal service to open the various opportunities to our emerging community. We would like to ensure that the infrastructure and challenges will not limit our efforts to deliver knowledge to our community.


Metrics and results overview

edit

This section provides metrics that will be used to measure each program's achievements conducted by Wikimedia Indonesia under this grant

Shared metrics

  1. Participants: The number of people who attend your events, programs or activities, either in person or virtually. This definition does not include people organizing activities, social media followers, donors, or others not participating directly.
  2. Newly registered: The number of participants that create new accounts on a Wikimedia project. These include users who register up to two weeks before the start of the event.
  3. Content pages: A content page is an article on Wikipedia, an item on Wikidata, a content page on Wikisource, an entry on Wiktionary, and a media file on Commons, etc. This metric captures the total number of content pages created or improved across all Wikimedia projects.

Grantee-defined metrics

  1. Active collaborations: the number of organisation or partners which actively collaborate at least 2 events during the grant period.
  2. Active community member: the number of active community member who spend their time to attend, organize or participate in the events, program, or activities, either in person or virtually with minimum 3 participation. This definition does not include social media followers, donors, or others not participating directly.


 
Details progress
Metrics Goals Achieved outcome Status Explanation
1. number of total participants 3,000 3,095
103.2% complete
The current situation makes our program can reach more participants comparing to pre-pandemic program that we focused on engaging community and newcomers in offline activities. This situation changes and shifts the perspective of online volunteering activities. People starts to engage more activities online as the situation limit the offline activities.
2. number of newly registered users 450 560
124.4% complete
Per previous explanation.
3. number of content pages created or improved, across all Wikimedia projects 61,500 38,702
62.9% complete
On target.
4. Active collaborations 15 3
20% complete
On track.
5. Active community members 100 148
148% complete
Per previous explanation.

Telling your program stories – all programs

edit

Community Development and Support

edit
Supporting More: From Five to Thirteen Communities

Last semester we saw a growing number of interests in forming new communities by several volunteers. With such interest, we managed to accommodate them and set up two forms for registering communities in May 2021. One form is aimed to re-register five communities already supported if they are willing to continue their communities. Another form is aimed for any group of volunteers that propose to us for community support. We assessed the responses to the latter form and decided which new communities to support. By July 2021, we started to support 13 communities, with 8 of them being new communities. The new communities include: Banjar, Batak Toba, Gorontalo, Madura, Mandailing, Medan, Nias, Wikisource Indonesia. Meanwhile, the five previously supported communities, i.e. Bandung, Denpasar, Jakarta, Padang, and Yogyakarta, continue to take our support.

Of all the supported communities, we divide them into three types: city-based, language-based, and project-based. The city-based communities include: Bandung, Denpasar, Jakarta, Medan, Padang, and Yogyakarta. The language-based communities include: Banjar, Batak Toba, Gorontalo, Madura, Mandailing, and Nias. Meanwhile, the project-based community consists of only the Wikisource Indonesia community.


 


Community meeting and social media support

edit

As mentioned above, the number of communities we support has increased from 5 to 13. The kind of support we give to these communities remains the same with the previous support, which is encouraging them to hold online monthly meetings. We provide for them a subsidy of phone credit worth IDR 50.000 for each meeting participant. The phone credit is purchased for them after their meeting reports are submitted to us. They may use the credit to buy internet packages to compensate for their data spent for online meetings.

Other than supporting their regular meetings, we also support them for their social media accounts. We also provide to them a simple graphic design app Canva that they can use to design their social media contents, along with a training session to introduce the app and how to use it. Most of the community designs on their social media are now made by using Canva.

Most communities have social media accounts in at least Facebook, Instagram, and Twitter. They make use of social media to introduce themselves to netizens and to outreach for new members. Since their activities vary across accounts, we regulate to what extent we can give our support on social media. We require each account to be active or post regularly at least once a week or four new and original posts each month and handled by one administrator. The administrator can claim their support at the end of each month after submitting a report to us. The support is a phone credit worth IDR 50.000 for one administrator of one account.

Internet package grant

edit
 

The Internet Package Grant still goes on from the previous period. Up to this time since the first half of 2020, this program has found a stable grounding. No rule adjustments have been made since the last report.

This grant supported volunteers by giving each of them a phone credit worth IDR 200.000 every month. Volunteers had to fill out our provided form in order to register. The period for volunteers to fill out the form was the first seven days for every month. We assessed their qualification based on two things, a volunteer must (1) have an account and (2) have made at least 100 edits in the last 30 days. After the assessment, the announcement would be delivered on the 7th day each month by email. The phone credit was purchased from the 8th day up to the 10th day each month for the qualified volunteers.

After receiving the grant, volunteers must edit 30.000 bytes from the 11th day until the last day of the month. They were allowed to edit only Wikimedia projects supported by Wikimedia Indonesia, i.e. Wikimedia projects in Indonesian languages. Volunteers passing the monthly task will be automatically included into the following month Internet Package Grant without reregistering or refilling out the form. They would be notified if they pass along through the regular announcement on the 7th day each month.

The average of grantees each month from July to December 2021 was 34.16 volunteers, increasing more than 11 points from the previous semester average (January to June 2021).

Book grant

edit

The Book Grant still also goes on from the previous period. This program has had a stable grounding since it was first launched in the first half of 2020, so there was not any rule adjustment since then.

This grant supported volunteers by giving a non-fiction book of their choice worth IDR 150.000 for every two months. Similar to the Internet Package Grant, volunteers had to fill out our provided form in order to register. Since this program ran bimonthly, we usually announced which month we opened the form. The period for volunteers to fill out the form was the first seven days of an opening month. We also assessed their qualification based on the two things, a volunteer must (1) have an account and (2) have made at least 100 edits in the last 30 days. After the assessment, the announcement was delivered on the 8th day of the opening month by email. Books of their choice were purchased by us from the 9th day up to the 11th day of the opening month for the qualified volunteers.

After receiving the book, volunteers must edit 30.000 bytes from the 1st day until the last day of the following month. They were allowed to edit only Wikipedia, especially of Indonesian languages, and submitted their edited articles to us through a page that we prepared on Meta. Volunteers passing the task can re-register for the next period of Book Grant.

The average of grantees has increased a little from the previous semester. If the previous semester saw 13 grantees from December 2020 to May 2021, this semester there were 14.3 grantees from June to November 2021.

Online scholarship

edit

The Online Scholarship is a continuation of the last semester program. Until this semester, the program has seen some adjustments from the previous semester.

Online Scholarship is a program to fund editing activities by individuals or groups for a period up to ten days. The fund is a scholarship per person for purchasing meals worth IDR 100.000 per day and a phone credit worth IDR 200.000. Each editing activity can be carried out by one to five people. Topics to edit must be related to Indonesia or vital topics. Edits must be carried out on any Wikimedia project of Indonesian languages, Wikidata, or Wikimedia Commons. Volunteers are free to choose how many days to do the activity up to 10 days and whether the days are sequential or not. With that time span, we expect that their activities should be finished in three months.

Each registrant had to have at least 10 global edits. Their proposals were submitted on Meta through a form that we have prepared. We assessed each registrant based on their number of global edits and their proposed activities if they meet the the allowed topics and Wikimedia projects. We announced the passed proposals by email. After that, each scholarship grantee must sign a letter of agreement. The scholarship was sent to their bank account after the agreement was signed. Ending their activity, each grantee must submit a report on Meta through a form that we have prepared.

Compared to the last semester that was held twice, the Online Scholarship was carried out once this semester, starting from July to September 2021. It is one of the adjustments that we made of this program since the steps to carry out this program are many while it needs careful supervision. Another adjustment is expanding the number of months from two months to three months. Even though it was held once, the number of granted proposals was the highest so far, amounting to 20 grantees.

Laptop and internet support

edit

Wikimedia Indonesia continued laptop and internet support for 5 Nias and Mandailing contributors. Each month the grantee could choose to create new articles, improve the existing articles, create or improve entries on Wiktionary. Per December 2021, there were 1.615 pages created.

From this activity we learned that not all grantees were actively editing each month. Support that lasted for a long time (1 year) did not guarantee that they would be consistent to work on Wikimedia projects. Therefore, for the next period we do not open this kind of support.

Lesson learned

edit

It has been quite a while that our communities never do in-person activities anymore due to the pandemic. Considering the Indonesian rate of vaccinations is quite promising and the number of cases are getting lower, we thought it would be great to reconsider allowing communities to hold in-person activities. Some of the community activities that we thought can be held in-person again are the community meeting and the gathering night. However, before they are actually taking place, careful planning needs to be done.


Education

edit
From participants to active volunteers

Although the COVID-19 pandemic situation in Indonesia has begun to improve since the government started to urge the vaccination program, we realized that we still needed to monitor the condition of the COVID-19 pandemic in Indonesia before deciding to hold offline activities (especially with the spread of the highly contagious Omicron variant). Therefore, in the first semester we chose to focus on improving the quality of our online programs. Even though the pandemic has hindered us from doing offline activities like we used to, we gained a significant thing from organizing online activities.

We were able to reach more new people by doing online activities, especially through our Online WikiLatih Course, which is a three-week Wikipedia editing course for the general audience. We have noticed that many of our alumni from Online WikiLatih Course have continued their contributions by joining our other programs, such as the program offered by the Community team, the Wikidata team, or the Competition team. Not only that they have also become active in joining our other programs, but some of them have also been actively participating in community activities. The number of involvements that we have gotten from our online program’s participants is certainly higher than the number or involvement we had gotten from our offline activities’ participants. We believe that it is because our Online WikiLatih Course is a program with such rigorous demands. Thus, the participants (particularly the ones who managed to complete it) might feel strong engagement with the activities itself, therefore they would be interested to partake in more events. This is clearly good news for Wikimedia Indonesia, and we hope that we will be able to maintain this, especially in terms of having more volunteers to join our program.


 

Online Meet-up

edit

Online Meet-up is still a routine agenda that we organize every month or every two months via Zoom meeting for 2 hours each. Participants who registered and attended this event are usually no more than twenty people. In every meeting, we edit various topics of Wikipedia articles, such as spices, flora and fauna, woman figures, and momentous events that occurred throughout the year.

In this semester, our communities have started to get more encouraged to organize their online meetups independently. As of December 2021, community meetups that have been successfully arranged include Balinese Wikipedia, Bandung, Medan, and Padang communities. Apart from editing Wikipedia in their respective local languages, the editing agenda in each community also sometimes involved adding new entries on Wiktionary, just like what the Padang and Bandung communities have previously done.

We hope that the Online meet-up program can provide a fun and safe space for our contributors to conduct editing activities together, as well as to exchange stories and experiences. It is an effective way for us to reach more contributors in different regions in Indonesia, since it is held online. However, maintaining the number of participants present at each meeting is still one of our challenges, especially because now we assume that people have started to get bored of joining online activities.

We offer WikiLatih in two different formats to have some flexibility when it comes to collaboration with partners. They are able to choose the best suitable format that would perfectly fit their needs and interests.

 

We managed to conduct 4 online courses until the end of the first semester. These online courses were our 6th to our 9th. When we first started this activity, we never thought that the attention and enthusiasm that we would get would be this massive. Recently, we had been getting more than 150 people registered for this activity, thus we have added more quota, from 40 people to 60 people for that reason.

From the 6th to the 9th batch, the numbers of people who managed to complete the course are as follows: (VI) 22 out of 42 participants successfully completed the course, (VII) 17 out of 39 participants successfully completed the course, (VIII) 29 out of 62 participants successfully completed the course, and (IX) 30 out of 61 participants successfully completed the course. These participants, who managed to finish the course, were given a certificate of completion by Wikimedia Indonesia. While the rest of them were participants who did not pass due to inactivity, resigned, and could not complete assignments until the course period ended will not receive that certificate.

The materials that were given in this online course were still the same. We also had regular online sessions with the participants that covered the topics that were not included in our main course. These topics included topics about our communities, Wikipedia behavioral guidelines, Wikipedia rules and policies, Wikimedia Commons, and Creative Commons License. Recently, we started to integrate the Welcoming Newcomer feature (Indonesian Wikipedia) in one of our online sessions for the participants to practice, so that they could make use of that feature to help them make some edits during their course period.

Live (WikiLatih Kilat)
edit

WikiLatih Live still becomes our communities’ go-to activity whenever they want to organize a WikiLatih, especially in their local languages. The format of the event is still the same. It is divided into two sessions, 2 hours each, by Zoom. There are usually 20-25 participants that join this event. As usual, during the sessions, we explain about Wikimedia, the movement, and the basics of editing on Wikipedia.

This semester, our communities were encouraged to hold their own WikiLatih Live for their local languages. Two volunteers from each community helped to carry out each of the events. One of the volunteers presented the materials, while the other one had to take care of the registration and the post-activity report. The training by the communities focused on the development of their local language Wikipedia, thus, they usually also used their local languages as the medium of instruction during the activity. This is what made this event quite different and more interesting. There were also quizzes at the end of each WikiLatih Live’s activity with prizes to appreciate the participants and to make them more motivated to participate as well as to contribute to our other activities.

From July to December 2021, WikiLatih Live has been held for the Mandailing, Medan, Sundanese, Javanese and Minangkabau communities. In the future, we hope that there will be more of our communities that would feel motivated to hold their own WikiLatih Live.

Advanced (WikiLatih Daring Mahir)
edit

Online Advanced WikiLatih is an online activity for contributors who have mastered basic Wikipedia editing skills but not enough to be able to make edits using codes. This can also include participants who have joined Wikipedia training activity before where they were equipped with basic editing skills.

For that reason, this training focuses more on source editing that is considered more “advanced” now since new editors nowadays tend to use the Visual Editor more than the source editing. Materials taught in this session include basic formatting using codes, templates, infoboxes, and image galleries. The session usually lasts for around 2 hours. An experienced Wikipedia editor will guide the participants throughout the session.

In the first semester this year, Online Advanced WikiLatih has been organized 2 times, which were our 6th, and our 7th. The participants were mostly alumni of the previous online courses.

We usually organize this event one or two weeks after an online course is completed or two sessions of WikiLatih Live are finished because we want the participants of those activities to join this event. However, we have had difficulty in finding experienced contributors who are willing to be a presenter at this event. We may have many experienced contributors, but there are not many of them who are willing to do this.

Mini WikiPelatih (Training of Trainers)
edit
 

Mini WikiPelatih 2021 is an activity whose main aim is to gather the active volunteers of Wikimedia Indonesia ⎼ both old and new ⎼ who are involved in the movement, so that they will be able to connect with each other by sharing their stories and experiences.

In this event, the participants were given some materials that we hoped would benefit them in the future. The first material was related to Open Educational Resources (OER) in which the presenter talked about what OER was, and the benefits of them, including the role of Wikimedia projects as one out of many Open Educational Resources. Our colleague from Creative Commons Indonesia helped us to deliver the material. The second material was about The Wikipedia Library (TWL) which explained about TWL and how Wikipedia editors could gain advantage from it. This was also aimed to encourage our new editors to increase their contributions so that they would be qualified to access TWL. The last material was regarding information literacy on the internet. We saw the urgency to equip volunteers, who were also Wikipedia contributors, with information literacy skills, in the midst of the massive amount of disinformation and hoaxes on the internet. We facilitated a group discussion and encouraged each group to do fact-checking with lateral reading techniques.

This semester’s Mini WikiPelatih was divided into two sessions. This was one of the things that we evaluated based on the evaluation form that we distributed to last year’s participants. Many of them were excited about the event, however they thought that it would be good if we could divide the event into two sessions. Having to sit down to join such an event for four and a half hours was not comfortable for the participants. Thus, we divided this event into two sessions, with a duration of 2,5 hours each.

We opened a registration form for this event and we were very happy to see the enthusiasm from the community members as well as the alumni of our Online WikiLatih Courses to join this event. We did a selection for the participants, and we invited them by e-mails. In total, we had 44 participants attending this event.

Even though the event went well and the attendees participated joyfully, we have not been able to measure the level of impact this activity had for the participants as well as our program in a more certain way.

WikiGuidebook (WikiPanduan)

edit

We started off this semester by reviewing the local language modules that had previously been translated. The modules are in Banjarese, Gorontalo language, Madurese, and Nias language. The modules had to be reviewed to ensure the quality of the translation, from the choice of words, the grammar, to the punctuation used in the modules. To achieve this, we involved four volunteers from each community to review the translation of the modules. The four variants of the modules have now been translated and are in the process of layouting and designing by the Communication team of Wikimedia Indonesia.

In addition to the local language modules, this semester we also managed to localize the three series of Reading Wikipedia in the Classroom modules by the Education Team of Wikimedia Foundation into Indonesian. We involved two active volunteers who were also Wikipedia contributors. The localization process took place from August to October. During the localization process, we were also in touch with the Education Team of WMF to make sure that we were on the right track with the localization. To monitor the localization process, there were meetings that we organized every week. We had to make sure that the modules would best suit the needs of the teachers or educators in Indonesia.

WikiWebinar

edit

We were aware of various misconceptions about Wikipedia in education, especially at the high school and college level. For this reason, we had the initiative to organize an online seminar entitled “Wikipedia's Misconceptions in Education” which is open to the public. We had 4 academics, as well as experts in their fields who were also active contributors to the Indonesian and local language Wikipedia to provide their perspectives on the misconceptions of Wikipedia that had been circulating around either in the society or in the Education field. One of them, who was an Associate Professor, happened to be a recent graduate of our online course.

The enthusiasm for this webinar was quite high. There were 447 people registered for this event and there were 173 participants who joined the event from various backgrounds: teachers, scholars, students, etc.

Office Hour (Bincang PDK-Komunitas)

edit

Office Hour, or what we usually call "Bincang PDK-Komunitas (Community-Education Talk)" is an attempt to provide our community members and volunteers a space for them to discuss anything with the Education team of Wikimedia Indonesia. They are allowed to ask questions or raise some suggestions for Wikimedia Indonesia. This activity is also held as a place for us to deliver information to the community members and volunteers. Office Hour is usually organized on Friday evening, every two months.

This semester’s Office Hours covered a few topics. The first topic was about the introduction to our Education program. Our community members and volunteers attended this session in which we had a productive discussion about the Education program for the next one year. The second topic was about partnership and collaboration. We gave our community members and volunteers an explanation of the steps that they would need to take if they ever wanted to have a partnership or collaboration with Wikimedia Indonesia. The third topic was about Welcoming Newcomers discussions for our local language Wikipedia communities. We invited Benoît(Trizek) from the Growth Team of Wikimedia Foundation to deliver a presentation about the Welcoming Newcomer feature that had recently been deployed on local language Wikis.

The three sessions of Office Hour are usually followed by Q&A, as well as discussion sessions. So far, the Office Hour sessions have been highly informative and interactive. We hope that in the following semester, we can continue to have Office Hour every two months in an attempt to communicate with our communities.

Online Talk Show (Gelar Wicara)

edit

Online Talk Show is an event in which we invite one to several volunteers to share their stories or experiences either live or recorded on our YouTube channel. We wish that by broadcasting it on YouTube, we are able to reach many more people. Since we started to go online, we have had several online talk shows, and this semester, we managed to have two online talk shows.

The first one was in July, where we invited 4 alumni of the Education internship program to share their stories when they did the internship, both offline and online. At that time, we also happened to open an internship opportunity for the next period. Therefore, we took advantage of this momentum to socialize the Education internship program. Then the second talk show we had was in August, where we invited the winner of the Best Newcomer Award of Wikimania 2021, Carma Citrawati from the Wikimedia Indonesia Community in Denpasar (Bali). She shared her experience of digitizing and translating Balinese lontar manuscripts and developing WikiPustaka.

For the next online talk show, we plan to also use different platforms such as Instagram, or Twitter so we can reach a more diverse audience. Apart from that, we also plan to take advantage of our own Instagram account, @pendidikanwmid, to do more promotions of our works and activities.

Education Internship (Magang Pendidikan)

edit
 


We decided to have another Education internship program this semester. The Education internship program lasted for 3 months, starting from September to December. During the internship period, the intern was not only involved in various Education program activities, but he also was mainly in charge of the module-reviewing process.

During his internship period, he authored articles about some educational activities that have been carried out for Wikimedia Indonesia’s blog. He was also required to join regular community meetings so that he could meet the people behind Wikimedia Indonesia’s communities and connect with them.

By the end of the internship period, the Education intern managed to fulfill his completion metrics. He compiled all his activities during the 3 months of the internship in a final internship report, then we evaluated the progress together. He also gave several recommendations, especially for the implementation of the next Education internship program.

The Education internship program is an effective means to attract more people to be involved in Wikimedia Indonesia activities in the long term. Therefore, we will organize it again in the next semester with a more well-prepared program planning.

Lesson learned

edit

Throughout this semester, we have organized activities that involve many more participants from various backgrounds. It has been a little bit more than a year that we have organized online activities. We certainly realized that we have gotten better in arranging online activities. We would not have had these improvements if our activities’ participants had never given us any feedback. We were fully aware that these feedbacks were important for us as they gave us an idea of the participants’ experiences during their activities with us.

As an example of this is the feedback that we received for one of our events last year, Mini WikiPelatih 2020. Almost all the participants filled out the feedback form that we handed out to them. The participants were very appreciative of the event, and the feedback was mostly positive, which humbled us. However, they pointed out a few things that we could improve for our future events which have similar schemes, such as the time management, the number of the sessions as well as each session’s duration that would be more effective and convenient for the next events.

Utilizing different platforms during online sessions could encourage participants’ engagement in the activities
edit

Something that is often overlooked when organizing online activities is the fun and enjoyable aspects of the activity itself. Arranging a fun and pleasant online event is not easy. Thus, to improve the participant's experience during our online sessions, we use free and lightweight learning support platforms.

We tried this once when we had Mini WikiPelatih 2021. We started using Google Jamboard and Padlet to support activities that required a moderate to large number of participants. The activities involved collaborative activities in which the participants could actively follow and jump in. They could collaborate by writing their opinions or things they would like to share directly on the Jamboard and Padlet.

Apparently, the participants got immersed with the activities they had to do but still in a delightful way. Therefore, we realized that integrating these kinds of platforms to our activities would certainly improve the experience of participants because they could involve themselves fully in doing the tasks.

GLAM & Partnership

edit
We welcome a new community and an intern!

In the first semester of APG 2021, GLAM Indonesia’s activities were still affected by Covid-19 pandemic. All activities, including workshops, seminars, and meetups, were forced to be held online, based on Wikimedia Indonesia’s official letter of notification. However, we were doing our best to engage with new kinds of activities. We highlighted four significant achievements during the first phase of our grant period:

  • Wikimedia Commons booklet translation entitled “Panduan berkontribusi konten di Wikimedia Commons” (A quick guide how to contribute your contents in Wikimedia Commons);
  • Map of GLAMs in Indonesia;
  • Establishment of Indonesian Wikisource community and proofreading competition run by the community to celebrate Wikisource’s birthday;
  • First GLAM’s internship at Wikimedia Indonesia.

Throughout this new year, we will continue to coordinate with GLAM institutions in the Netherlands through Wikimedia Nederland, for collaboration on collection donations, metadata, structured data, and dissemination of collections related to Indonesia. Until the end of 2021, there were six institutions involved in this early collaboration. Some of them are still exploring the cooperation and digitizing and the rest have already supplied access. For the whole of first semester of 2022, GLAM Indonesian team will continue to expand our collaboration with GLAMs, such as conducting #1lib1ref campaign with libraries, engaging with community to run Wikisource contest, and work together with other entities to conduct a series of events to commemorate Public Domain Day 2022 in Indonesia.


 


Cooperation with GLAM Nederland is the first agenda of GLAM Indonesia at the beginning of the first semester of GLAM 2021. Continuing the previous successful collaboration on WikiSejarah, it was deemed necessary to continue the donation and the improvements of the GLAM Nederland’s collection related to Indonesia via Wikimedia Nederland. Activities in this collaboration will open collections related to Indonesia in six GLAM institutions in the Netherlands, translate Dutch metadata into Indonesian, add structured data and use the collection on the Wikimedia project in Indonesia. Not only that, it has been planned for every institution to have a talk show, to promote and to discuss the collections that have been opened. This collaboration and communication will run throughout GLAM 2021 and is likely to continue, with some institutions involved and the number of collections that have not been digitized. The institutions involved are:

The opened collections will certainly be valuable and become research material for researchers as well as for content creators. Metadata and structured data in Indonesian make collections appear in searches.

In 2021, the booklet "Panduan berkontribusi konten di Wikimedia Commons" was published for those who love photography and are still confused about how to contribute to Wikimedia Commons. The existence of the Commons platform is certainly different from other existing photo platforms, both in terms of appearance, menus, features and of course anything related to copyright. All are discussed in detail in this booklet. This booklet is an adaptation of the booklet entitled "A guide to contributing content to Wikimedia Commons". There were some additions made by Rachmat as translator and author. The booklet is printed for physical dissemination and can also be downloaded at Wikimedia Commons.

GLAM Indonesia as a program that digitizes Galleries, Libraries, Archives and Museums certainly needs to have complete data about these institutions. Therefore, the GLAM team together with the Wikidata team took the initiative to make a GLAM Indonesia map. This data will serve as an overview of the data to see institutions that have opened their collections on the Commons, their articles on Wikipedia and of course provide an overview of our GLAM institutions in Indonesia. The GLAM Indonesia map can be accessed on GLAM Indonesia website and on Wikidata.

Still with something new at GLAM 2021. It is from a new community which is the first and only project-based community, Indonesian Wikisource Community. This community is active in conducting online meetings and reading-testing together. Not only that, the Indonesian Wikisource Community on the anniversary of the Indonesian language Wikisource held a reading test competition. The competition invited many contributors participating on the Indonesian Wikisource site. This activity will be on the community agenda and the community will also be the committee in the future Wikisource competition, both for Indonesian and regional languages. Meanwhile, the Indonesian Wikisource Workshop was carried out together with the Banjarmasin and Gorontalo Wikimedia Community.

GLAM Indonesia provides internship project for the first time. The internship in GLAM Indonesia in 2021 was the first internship program. This program was immensely helpful in the digitization process and other matters related to GLAM Indonesia activities. The need for the interns this time is to help the cooperation that is being established between GLAM Indonesia and GLAM Netherlands. It was Ni Putu Diah Ashanty who was selected by GLAM Indonesia team from among twenty applicants, who came forward as an intern. Graduated from Dutch Studies at the University of Indonesia and armed with experience translating the De Javasche Bank collection, Diah did translations in several categories on Wikimedia Commons which were approved by GLAM Netherlands. The apprenticeship will be part of GLAM's permanent program which will then focus on digitization.

Meanwhile, regular activities carried out this semester were the 1lib1ref workshop and the ISA Campaign., 1lib1ref workshops have been held with a number of communities and institutions including; Pustakawan Mendunia and Muhammadiyah University of North Sumatra. The ISA campaign was held twice; ISA of Wiki Kaleidoscope and ISA of Siboga Expedition.

Lesson learned

edit

GLAM activities in 2021, which are dominated by online activities which have been carried out for almost two years. They were certainly not as effective as offline activities. The interest in joining online activities seems to be boring for volunteers. The same vibe was also found in collaboration with the GLAM institution, which preferred to be held offline. GLAM Indonesia needs to conduct new activities, to ensure GLAM's goals are achieved, by helping GLAM institutions share their cultural resources and knowledge with the world through Wikimedia projects, one of which is Wikipedia.

However, of course, we also received numerous inputs for this program, such as requests to resume working with new institutions and create new, different events targeting GLAM institutions. This is still a limitation for us because we are still adapting to the current pandemic status that is unpredictable day by day, and still work from home. Then, there are inputs for the future. The Wikisource competition is not only for Indonesian language Wikisource, but also for local language Wikisources.


Content Enrichment

edit
Our commitment to digital literacy in Indonesia

At the end of 2021, Wikimedia Indonesia organized Tantangan Hak Digital (Digital Rights Challenge) in collaboration with EngageMedia, a nonprofit organization focused on digital rights advocacy. We make this challenge because of limited understanding of digital literacy and rights in Indonesia, e.g., user data protection, laws, and their rights in cyberspace, and cybercrime. Much of these articles are just stub articles or unavailable in Indonesian Wikipedia, and it’s necessary to expand or create these articles for informational purposes to the Indonesian citizens.

As the name suggests, the topic for this challenge is digital rights. They covered various topics from digital rights management, identity theft, to freedom of expression. This challenge was organized for one month, from 8 November to 8 December 2021. This event resulted in 65 articles being improved and 35 new articles being created from 11 participants. All of the participants got digital vouchers and exclusive merchandise. After this challenge, we hope people will better understand digital literacy and their digital rights.

Jatayu project

edit
 


For nearly two years Covid-19 pandemic has happened in Indonesia, there have been significant changes in our way of living. These changes include financial management during uncertainties, processing all available information to avoid misinformation or disinformation, and physical or mental problems during recent situations. These topics are not well covered in Indonesian Wikipedia.

That is why we organized a new writing competition, Proyek Jatayu. This competition runs for five months, from November 2021 to March 2022. Each month, there are different topics, ranging from financial technology, critical thinking, primary needs, physical health, and mental well-being.

In this competition, we are using a system similar to WikiJelajah 2.0, i.e., a points system. Participants may pick to expand an existing article or create a new article. We changed some of the rules. For example, participants may also choose to translate articles from other languages. In addition, we are revising how to calculate the score for the grand prize. Proyek Jatayu emphasizes the good articles reviewed by judges, accounting for 60% of the score weight. The other 40% is from the accumulation of monthly points.

WikiKaleidoskop

edit
 


WikiKaleidoskop, as a photo competition with a different concept from the previous photo competition, is one of Wikimedia Indonesia's methods to add photos licensed under CC BY SA 4.0 on Wikimedia Commons. The outcome photos from participants can visualize articles on Wikipedia to make them easier to understand and more attractive. Not simply that, this competition can be an opportunity for photographers to increase their ability to take photos of various objects.

The WikiKaleidoskop photo competition starts in January 2021 and ends in December 2021. The themes during the second half of 2021 are:

  • Herbs and seasoning.
  • Tangible and intangible cultural heritage.
  • Tools and equipment.
  • Traditional buildings.
  • Health and pandemics.
  • Activities and professions.

The choice of quite varied themes has proven to be able to reach many people so that the number of participants is increasing every month. The number of participants increased to 1.559 people. Participants are primarily new users of Wikimedia Commons. Some participants only participate in the competition in certain months, but some consistently participate every month. From January-December 2021, the WikiKaleidoskop competition managed to get 3.344 photos consisting of 12 different themes. In addition to the increasing number of participants, we found that 460 from 3.344 photos were used for 590 articles in various Wikimedia projects such as the Indonesian Wikipedia, the Javanese Wikipedia, the Russian Wikipedia, and Wikidata.

Lesson learned and next step
edit

During November-December 2021, we had two events with similar format runs near-simultaneously, Proyek Jatayu and Digital Rights Challenge, and impacted the number of participants. In the Digital Rights Challenge, we targeted more than 25 participants. However, only 11 participants contributed to the challenge, which is even below from participants that got prizes, i.e., 15 participants. We suggest two competitions with similar formats (e.g.writing competition) must be organized at different times. Two competitions may run at the same time if they have different formats (e.g., writing competition and photography competition)

Using points format that is similar to WikiJelajah 2.0 and different topics each month in the writing competition makes participants more likely to improve old articles than create new ones. We believe that is a good thing since old articles that covered fundamental things are relatively untouched during regular times are now more comprehensive than before. However, this format has some drawbacks. For example, many participants submit their articles at the end of the month. We have time constraints for checking their articles because we need to announce monthly winners as soon as possible. So we may not be able to check their articles thoroughly. The drawbacks also include "loopholes" seeking participants and using unreliable references, though this occurrence is relatively lower than the previous competition.

Another problem was when participants translated an article from English Wikipedia to Indonesian Wikipedia. Many of the translated articles are poorly written and only use machine translation without further editing. We will reconsider article translation rules for the next competition.

From our experience during the WikiKaleidoskop photo contest, some of the newcomer problems still happened. For example, unregistered participants tried to upload their photos for the competition, some of the photo descriptions were not written correctly, or uploaded photos in the wrong period. There is a thought-provoking finding; most of the monthly top ten photos come from a few participants. We assume most professional photographers are reluctant to participate in this competition because they have concerns about copyrights. This problem has become our challenge to advocate open culture and its benefits.

 


We held all of our Wikidata events online during the second half of 2021 with the pandemic still ongoing in Indonesia. Despite that, in line with our visions, we continued our steps to develop the Wikidata community in Indonesia and to improve usage and quality of data regarding Indonesia in Wikidata. With no offline events planned in mind up to the end of the year, we focused on increasing the quality of our online events and our methods to attract and interact with the Indonesian editors and our volunteers through online means.

Attracting new editors and involving them and our current active ones as volunteers in our events are important in keeping the emerging community to develop further and in making an Indonesian representation, both in terms of data contents and editors, among the whole Wikidata community as a multilingual project in Wikimedia. However, in involving our volunteers we still encountered some problems both regarding their numbers as well as the opportunity each of them have and their interest in participating in our events. In this period, we managed to hold 11 events on Wikidata which consisted of 2 Wikidata trainings, 4 online meetups, 2 workshop events, and 3 datathons. We chiefly used Indonesian topics for the materials of our events. These are ranging from Indonesian islands and regions, to notable Indonesians and Indonesian culture. Using local topics not only will on itself improve the quality and the availability of data regarding Indonesia in Wikidata, but will also make our events more relevant and relatable to our editors.

Events and activities
edit
Engaging students with Wikidata via data visualization competition

From July to August 2021, we collaborated with IT Del (Del Institute of Technology) to hold an internal event for data visualization competition. The event attracted 23 teams of two and 16 submissions. Their projects were presented via Google Sites, from Indonesian traditional food to Indonesian museum. The winner is PawangCORONA team who visualize the Indonesian hospital data, they used Wikidata Query Service to analyze the list of Indonesian hospital, beds number, and adding a short information from their respective Wikipedia page. Another prominent winner are Mugiwara with Indonesian Nature Statistics, Elysian with Indonesian Province Fun Fact, and Yeagerist with their Indonesian National Hero Data. It was a lot of fun for student to play with the real data and visualizing their imagination.

Our Wikidata training is a little bit different from the online training we conducted in the previous periods, where now we have our volunteers participating as the training coordinators. Both of our events are held online and our previous experience was that we could not find enough participants to attend the event. In overcoming this, we experimented with opening our registration for a longer period—in some occasions until the morning of the day of the event—which in turn allowed more people to register. As not all of the people who registered would eventually participate and attend the event, this has actually allowed more participants to attend the event rather than using a predetermined quota on the registration form and closing it once we had the number. New methods on giving the training material were also conducted where instead of teaching on several backgrounds regarding data and only showing the participants a live step-by-step example in editing Wikidata, we emphasized more on the structures of Wikidata and a concise guide to edit Wikidata alongside a live demonstration. Using these methods, we managed to keep the duration of the event on time as planned at around 2 hours, less than the previous average of 2.5-3 hours, while still delivering the important parts of our course. We used the Zoom platform in our training. In some events, participants who were keen to ask a more advanced question could do so by staying on Zoom after the event had completed. This way, the focus of the events was still in the material we prepared on the basics of Wikidata.

We held 2 training events which were open for public events aimed at beginners in Wikidata or Wikimedia projects. Our workshop events were held similarly. We held 2 workshop events for college students in partnership with the Computer Engineering Laboratory, Faculty of Engineering, University of Lampung. Although we are on track with our workshops, we are still below our target for the number of training sessions. We hoped to involve as coordinators our regular participants in our previous events where they could prepare the materials for the participants, helping the participants, and being the contact person for the event. Our plan was also to try to involve different persons in each of our events so each of them at least gets to be a coordinator once in this period. However, in planning some of our events, we learned that not all of them have the time and opportunity while some others were not interested. In composing our material, we also encountered limitations with the sources we use such as the official Indonesian intangible culture directory which on some occasions was not accessible or with the item itself being already present in Wikidata.

The online meetups in this period were held more frequently with 4 events held so far. In the events, we invite the Indonesian Wikidata editors to meet with each other online using our Zoom platform where they could share their experiences, ask questions, and talk to each other on general Wikidata topics while also editing Wikidata. In each of the meetups we also involved a volunteer as the event coordinator similar to those in Wikidata training and workshops. The meetups were more popular than during the previous period and we received more participants and more engagement from them as the events became more regular and participants became less hesitant to start a conversation. These meetups were more informal and, in their edits, participants were using a material we prepared beforehand on various Indonesian topics. Through these events we also introduced them to Wikidata gadgets where in each event we introduced a different gadget to help in their editing or the ones adding a new feature on Wikidata interface. Participants were enthusiastic in trying the new gadgets and overall having conversations around Wikidata and Wikimedia projects.

We continued our data edit-a-thons (datathons) to its second year where we held 3 datathon events in this period. Our first two datathons were where participants’ contributions were judged only in a special predefined topic of items. We used the Outreach Dashboard article scoped feature to limit the query done to the participants’ contributions based on a PetScan query. Outreach Dashboard then would only count those contributions to the specific list of items resulted by PetScan. This was a different approach in judging where we previously would query the items edited by participants and remove the ones that didn’t match the criteria. It also enables us to conduct future datathons using a specific topic. The third datathon was our special datathon during the week around New Year’s Eve and was held up until January 2022 and similar to the new year datathon we conducted before, this one did not have a predefined topic of items. Our new year datathons were held on a special occasion so we would want our participants to contribute in the topic they “specialized” more. However, we also had our Wikidata anniversary datathon but in this event, we focused on data using the Indonesia related-properties. Four Indonesian properties were created during the period of July-December 2021, adding to the total of 22 properties. Several difficulties arose during the judging process where some participants created a lot of items regarding subjects already present in Wikidata (such as geographic features) and we had to correct all of them virtually one by one. This had also resulted in the delayed announcement of the winners of this particular event. We had encountered the issue regarding duplicate items about geographic features before however at that time we regarded it as a minor issue as not a lot of participants were editing the topic.

All of our events were promoted through our social media accounts and message groups. We maintained a Twitter account where we also held bi-weekly quizzes. On each quiz after the answer to the question was revealed, we accompanied them with some information about the subject on various topics on Wikidata including but not limited to Wikidata constraints, queries, and properties and how to use them. We used a rule where a winner could only win once in every 6 months period so after winning they were not allowed to win or try to answer the next quizzes. This was put in place to prevent audiences with more advanced knowledge of Wikidata taking advantage over others who did not have the same advanced knowledge and we would also end up with a lot more people able to participate. Some examples of the quiz announcements are here below:

Lesson learned
edit

Continuing our events with more frequent activities meant that our community is continuing the achievements they have made during the last year. Using locally relevant sources for the materials of our events are important and a reliable supply of sources is needed in the long run. We also have to adjust the rules of our datathon to minimize the issues arising from mistakes in editing. As we are still planning to involve our community members as coordinators in our events, we plan to hold a training of trainers (ToT) event in the upcoming period. We are also watching the situation around the pandemic and whether offline events are possible in the near future.

Javanese OCR

edit
Literature study and experiments to improve OCR system capability

Throughout this semester, the TRAWACA team focused on conducting a literature study on techniques that could be applied to improve the capabilities of the previously built Javanese script OCR (OCR Cakra). The topics that have been explored and implemented by the TRAWACA team are techniques for improving segmentation and feature extraction processes, techniques for scaling the characters, and techniques for rotating the characters.

Literature Study for Improvement of Segmentation and Feature Extraction Process
edit

The problems faced by the TRAWACA team in the previous year's project (APG20), were over-segmentation and under-segmentation. Therefore, the literature study in this semester is focused on overcoming this problem, as well as exploring options for feature improvements in order to increase the accuracy of character recognition. The selection of the right segmentation method and representative features, both are closely related to get good character recognition results.

At first the TRAWACA team reopened a collection of libraries that had been read about the OCR process, segmentation, feature selection, pre-processing and others related to character recognition. Regarding segmentation, it can be concluded that character or character segmentation can be done using 3 approaches:

  1. Segmentation by prioritizing the final result (cutting characters).
  2. Segmentation with contextual postprocessing.
  3. Recognition-based segmentation.

According to the TRAWACA team, "thinning" can be used to equalize the thickness of characters from various document resolutions, which until the previous stage caused accuracy problems. In Javanese OCR, the TRAWACA team will use thinning in the feature extraction process.

The water reservoir techniques might be used to overcome under segmentation provided that it was not entered directly in the Java OCR algorithm that already exists, but must be reviewed under certain conditions before being implemented, for example detecting whether the segment is really under segmented or not. This manuscript simulates the connection of two characters that should be separated as a concave container of water, in order to find the exact location where to cut the connection.

Because the literature review for the character segmentation process has not provided an appropriate solution in the case of Javanese OCR, for the time being the TRAWACA team has decided that in the next quarter it will sharpen the algorithm that has been proposed in APG19, namely the “clothesline” or “clothesline” algorithm based on Connected Component. Labeling is continued by merging for elements that are still considered part of the same segment.

Meanwhile, for the selection of features that are more representative for the Javanese script, the TRAWACA team found some inspiration from Thai script recognition texts, for example "Multi Feature Extraction For Printed Thai Character Recognition" by Kawtrakul and Waewsawangwong (2000) which uses character contours in the form of writing directions. in each zone, then peripheral information is the edge distance of the bounding box or the closest letter segment box to the character for each zone, which indirectly represents the writing contour.

Another inspiring manuscript is “Thai Handwritten Character Recognition by Genetic Algorithm” by Pornpanomchai, et al. (2011) which extracted the presence and position of vertical, horizontal, zig-zag lines, tails (ends of lines), loops (holes) and the number of loops and then poured them as a series of “DNA” characters in the form of a series of genetic descriptors as their feature vectors.

Thus, the two strongest candidates to be implemented into Cakra OCR are contour features and genetic descriptors.

Implementation of Scale Conversion for Characters
edit

The rescaling of the character image is carried out using the same technique as last year. The image scale is converted into pixel dimensions [118,125] with dimensions [height, width] which is 1.5 of the average scale of all characters. The difference lies in the storage process. This time the results of the scale changes are stored permanently on the hard disk.

By saving the image results of characters that have the same size, several labeling inconsistencies were found between the data collected in 2019 and 2020. Thus manual repair was carried out again. Most of the labeling errors occurred in 2019 data which was previously annotated by crowdsourcing. Labeling errors in training data from 2021 were also found but in very small numbers because the annotation process was carried out by Javanese experts. In this case, it seems that the process of improving annotation labels is endless and tiring, but this process must be carried out considering that it will affect the training process and model learning which will affect the precision in character recognition in the test data later. The types of errors encountered when scaling and saving characters are wrong class and misreading of the characters.

Correction of label errors is carried out on the original data, so that when other data augmentation techniques are applied, it will minimize repeated errors. For damaged characters that are still labeled, then deletion is carried out if the number of characters that are the same as the damaged ones exceeds 50, if there are only 2 or 3 samples of characters in a class, and one of them has noise or less significant damage, then the characters are retained. To maintain the consistency of labeling quality, the initial output of the scale change process was deleted, and a reprocessing of the character scale change was carried out on the improved training data.

Implementation of Rotation Techniques for Characters
edit

One of the data augmentation techniques is to rotate or rotate the image with a tilt angle. This technique was chosen because of its simplicity and for computers, a character, for example ꦮ 'wa' letter which is perpendicular and when the similar 'wa' is tilted to the right or to the left at an angle of 2 or 5 degrees, is considered a different character. Initially, the rotation process was carried out using the CV2 library from OpenCV. When displaying (printing) the contents of an array containing class label names with a sample size of less than or equal to 50, an incorrect annotation was found with a unique label naming. The solution is to write the correct label in the training data. original and training data from the re-scaled results.

The problem of truncated parts of the characters is solved by learning some techniques and functions that will not cut the image of the characters when rotated. Thus the cv2 library is deprecated, but uses the cuteils library with the rotate_bound() function. The pre-rotation stage is the same as the one used when rotating with the cv2 library, but the rotation process uses a different function.

There is a problem with this function, it makes the unknown areas, namely areas outside the ROI (Region of Interest) are assigned a value of 0 so that they become black. This can be overcome by using the constant interpolation technique, namely replacing the pixel value of the area outside the ROI with a gray degree value above the set threshold value so that when the character image is read later (thresholding process) the value will be changed to 255 or white color.

As explained earlier, the rotation process is only applied to character classes with a sample size of less than or equal to 50. Unfortunately, out of the 851 classes, 762 classes have a sample number of less than or equal to 50. The rotation process is applied to each class sample instance, so the total number of results of this rotation is very large. However, the increase in the number of samples still does not solve the case of inequality in the number of samples in each class.

Lesson learned
edit

The problems faced in this repair stage are more complex than starting the creation of the Java CAKRA OCR. The literature sources regarding segmentation and feature extraction that are explored are mostly concerned with training and testing materials that have been prepared according to the problem constraints, so that from the temporary analysis, the results have not been tested for application in general conditions (not prepared). However, these libraries provide ideas to overcome the problems encountered in the feature extraction process, where the current constraints are not only poor segmentation but also non-uniform document image resolution.

The segmentation method suggested by the literature that has been studied solves more specific problems, while from the results of testing during the APG19 and APG20 projects, Java OCR found both over-segmentation and under-segmentation. Adding a specific method to the algorithm that already has CAKRA's Java OCR analyzed can cause bleeding or conflicting rules. Therefore, the TRAWACA team decided to refine the algorithm that has been proposed in APG19, namely the “clothesline” algorithm that utilizes Connected Component Labeling (CCL) followed by merging rules between segments that allow them to be combined into a character segment. This sharpening will be carried out in the second quarter. Roughly speaking, this algorithm has succeeded in isolating non-text elements from documents as in the report on separating image elements from documents in APG19. Statistical analysis may be applied to the segmentation results to detect under-segmentation, and then apply specific methods to overcome the problem.


 

Financial report

edit

Revenues received during this six-month period

edit
  • On 3 June 2021, WMF approved WMID grants amount of USD 327,467.
Conversion rate

Rp 1 → US$ 0.000069

Note

edit
  • S1 is per 31 December 2021

Another format

edit
Revenue Source Currency Anticipated S1 S2 Cummulative Explanation of variances from plan
FDC APG Grant IDR 327,467.00 229,226.90 229,226.90 Initially, our proposal budget was USD 355,162, but because of the pandemic and after consulted to WMF Grants Team, we decreased our budget and assumed that on 1st October 2020 the situation would be normal.

Spending during this six-month period

edit
Expense Currency Budgeted S1 S2 Cummulative Percentage spent to date Explanation of variances from plan
Office and admin support USD 30,851 9,277 9,277 30.07%
Salary USD 150,073 70,126 70,126 46.73%
Community Development USD 26,215 10,000 10,000 38.15%
Community Support USD 29,291 0 0 0.00% For our offline conference in the second semester
Education USD 14,638 5,731 5,731 39.15%
GLAM and Partnership USD 8,368 1,402 1,402 16.76% Will have more offline activities in the second semester
Wikidata USD 11,585 2,578 2,578 22.26%
Javanese OCR USD 5,003 690 690 13.79% Will have more offline activities in the second semester
Competition USD 15,001 3,193 3,193 21.29% Most of the competition prizes will be buy in the second semester
Program Support USD 36,443 12,742 12,742 34.96%
Total USD 327,467 115,740 115,740 35.34%

Compliance

edit

Is your organization compliant with the terms outlined in the grant agreement?

edit

As required in the grant agreement, please report any deviations from your grant proposal here. Note that, among other things, any changes must be consistent with our WMF mission, must be for charitable purposes as defined in the grant agreement, and must otherwise comply with the grant agreement.

Are you in compliance with all applicable laws and regulations as outlined in the grant agreement? Please answer "Yes" or "No".

  • Yes

Are you in compliance with provisions of the United States Internal Revenue Code (“Code”), and with relevant tax laws and regulations restricting the use of the Grant funds as outlined in the grant agreement? Please answer "Yes" or "No".

  • Yes

Signature

edit

"All activities conducted by Wikimedia Indonesia were and are consistent with its status under Internal Revenue Code section which remains unchanged since the proposal and related materials were submitted to Tides Advocacy. Wikimedia Indonesia warrants that it is in full compliance with its grant letter dated 06/11/2021 and that, if the grant was subject to any restrictions, all such restrictions were observed."