Grants:Programs/Wikimedia Community Fund/Rapid Fund/WikiTermBase 2.0 (ID: 23553206)/Final Report
Application type: Tech project
Parts 1-3: Project and impact
1. What was built or achieved during the project, and how did it align with your original goals, milestones and technical plan? (required)
- Building a Morpho-syntactic parser (100% achieved): We built a fully functional Arabic-native morphosyntactic parser with open-source code that’s directly integrated into Arabic Wiktionary through a comprehensive set of templates {{تصريف/ Derivation]}}. The Python script is being actively revised with linguists to increase coverage, currently covering the derivation patterns of +95% of Arabic language verbs (we track coverage through the statistical distribution of Arabic verb types). This parser’s integration with Wiktionary means that +95% of Arabic Wiktionary pages can now be standardized and automatically populated with full morphological forms using this template, which amounts to at least two dozen derivations per verb in Arabic.
- Extra goal: Semantic Synthesis tool (100% achieved): During development, we decided to build an entirely new feature and UI of WikiTermBase that takes it a step further in full-integration into Wiktionary: a Semantic Synthesizer. This tool drafts the semantic descriptions or glosses of Arabic verbs from pre-selected secondary sources (other dictionaries), then adds them into a fully-structured Wikitonary draft ready with the word’s entire morphological data (see above). The draft page is then reviewed by Wiktionary editors before approval, and has been endorsed by lexicographers from Damascus Academy of Arabic Language upon thorough review. The tool is live on Toolforge, and has received great community feedback.
- Default Arabic Wikipedia gadget (partly achieved): While we significantly developed our tool and added new features to enable full integration with Wiktionary and maximum automation of automating lexicographic data addition, these new features introduce additional development overhead. We believe our tool makes a strong case for a default tool now not only on Arabic Wikipedia (for term standardization), but also on Wiktionary (for page drafting and creation). We also consistently increased our tool’s user base and adoption, now reaching a stable +50 active users and +400 total users, thus ranking as top 40 among 120 non-default tools, in spite of being a new addition that’s only 1 year old. We plan on launching a proposal to the community to move ahead with this default function once we’re ready, by EOY 2026, but we don’t wish to rush this critical process just due to the grant timeline.
2. Share links that demonstrate your project's impact, usage, and technical outcomes. (required)
Required links:
- Project page on relevant Wikimedia spaces (e.g. Phabricator, Wikimedia projects, Toolforge)
- Code repository (e.g. Gerrit, GitHub or GitLab)
- Documentation or user guides
- Dashboards, metrics tools, or analytics used to track usage or contributions
Optional links you may include:
- Diff or mailing list announcements
- Community feedback
- Demos or product presentations
- Survey results or user testing feedback
- Examples of integrations or usage within Wikimedia projects
- Project page on Wikipedia
- Improved existing tool on ToolForge: WikiTermBase
- Developed brand new tool on ToolForge: Arabic Lexemes Senses Review
- Developed templates on Wiktionary: {{تصريف}} (automated morphosynctatic parser of Arabic verbs), {{معجمية}} (full integration with Wikidata’s lexicographic data)
- Code repository on Github: WikiTermBase
- Arabic Lexemes User Guide on Wiktionary
- Tool Usage Statistics on Arabic Wikipedia: Showing our tool ranking within top 50%
3. What are the key lessons you learned during this project, both technical and non-technical? (required)
- Tool development is an adaptive process. While we initially thought the main feature we lacked was morphological parsing to fully integrate our tool with Wikidata and Wiktionary (thus making it a true part of the Wikimedia ecosystem and a good default tool), we quickly realized that editors wanted to use it to add new words and sense as well. We pivoted and developed a brand new sub-tool: the Arabic Lexemes Senses Review. With this new tool, we expanded our scope from a multilingual term base into a fully-blown lexical tool that automatically parses +25 Arabic morphological forms for +95% of verbs, synthesizes word definitions from trusted and public domain secondary sources, and ranks them by foreign language equivalence according to +50 dictionaries to enable editors to both standardize Wikipedia terms and simultaneously fill a much needed void in evolving Arabic Wiktionary.
- User adoption is built on trust and word-of-mouth. While our tool covers a community need that we long confirmed and has shown productivity gains of 10-20% with tested users, adoption has been most driven by the Arabic editor community familiarizing themselves with the tool and talking about it to each other. This is why we’re increasingly involving the community in testing and feedback loops.
- User adoption will hit a limit over time. We’ve seen active users fluctuating between 40 to 60 each month, with little room for increase in spite of extensive technical improvements. We believe that making the tool default “opt-out” instead of “opt-in” will significantly drive these numbers and boost editor productivity, but we don’t want to rush this process before the tool is technically ready and all new features have been tested for a few more months.
4. How did the Wikimedia community or your target audience engage with your project during its development or release? (required)
We engaged the communities of both Arabic Wikipedia and Wiktionary extensively throughout development:
● User-testing: Consulted with half a dozen testing users of relevant expertise during the development phase of the Arabic Lexemes tool (diffs: user 1, user 2, user 3, user 4, user 5, user 6).
● Community village pump: Extensively engaged the Arabic Wiktionary community in a village pump discussion for tool feedback and for feedback on morphosyntactic parsing, amounting to two dozen back and forth responses engaging Wiktionary admins and top editors.
● Live workshop: Organized an introductory tool walkthrough interactive session with 84 registered attendees.
5. What risks or challenges did you encounter (related to delivery, safety, or security), and how did you address them? (required)
- Linguistic faithfulness: Our project builds complex linguistic tools with high stakes and impact, a significant challenge was ensuring the accuracy and reliability of our parsers and synthesis tools. To address this, we directly involved linguistic experts and lexicographers as consultants, including Mamoun Al-Saghirji (member of Damascus Academy of Arabic Language), as well as Yahya Mir Alam (Professor of Arabic Morphology and Syntax and member of Mecca’s Arabic Language Academy).
- Community expectations versus technical capacity: Our project is highly technical, so meeting community expectations and requirements for the tool’s acceptance was a major issue. Thankfully, our incredible software engineer single handedly built an entire human-in-the-loop UI to ensure editors had the chance to review the tool’s output and act as approvers before we passed any content into Wiktionary and Wikidata.
6. Who will maintain the project going forward, and what is your plan for long-term maintenance? (required)
The project’s volunteer team is strongly committed and motivated for continuous oversight and maintenance of this tool. Our team has been working together now for nearly three continuous years non-stop to deliver the tool in its current form. While we highly benefited from a Wikimedia Foundation Strategy Grant (11K USD) and this Rapid Grant (5K USD), we have also managed to secure independent funding from UAE’s Literature Foundation (10K USD), and continue to move forward with development.
(questions 7-9 are skipped)
Part 4: Financial reporting
[edit]10. Please state the total amount spent in your local currency. (required)
4253.75
11. Please state the total amount spent in US dollars. (required)
4893.4
12. Report the funds spent in the currency of your fund. (required)
Provide the link to the financial report https://docs.google.com/spreadsheets/d/1SB9mQrHjQIDmdHYNZh5qbd50M7Dgp1SA7XJzytjL-Ko/edit?gid=928458698#gid=928458698
12.2. If you have not already done so in your financial spending report, please provide information on changes in the budget in relation to your original proposal. (optional)
1. We increased our computing budget (Claude Code) from $100 to $1,000. This change was the result of an entirely new tool development that took place halfway through the grant period, during which our software developer was able to create a brand new semantic synthesis tool that drafts full Wikitionary structured pages with morphological, syntactic, and semantic data using a mix of secondary sources. The drafts are reviewed and submitted by tenured editors through a fully built-out interface. Because the tool received high praise during testing with expert lexicographers and linguists, we decided to highly invest in this tool's compute, which involves significant LLM calls. 2. Staffing expense changes: Our Software Developer voluntarily decided to make his participation in this project phase pro bono, therefore saving about 2,075 USD in software development costs that we originally budgeted for. In addition to the above, we shifted the rest of this surplus to both our linguists (up from 1,150 USD to about 1,650 USD) to cover increased need for reviewing machine-learning outputs, as well as to our Project Manager (up from about 845 USD to nearly 1420 USD), due to the increased need for supervision and tool testing ops. 3. Dictionaries required slightly more budget than we originally planned for due to high shipping costs, therefore going up from 235 USD to 360 USD.
13. Do you have any unspent funds from the Fund?
No
13.1. Please list the amount and currency you did not use and explain why.
N/A
13.2. What are you planning to do with the underspent funds?
N/A
13.3. Please provide details of hope to spend these funds.
N/A
14.1. Are you in compliance with the terms outlined in the fund agreement?
Yes
14.2. Are you in compliance with all applicable laws and regulations as outlined in the grant agreement?
Yes
14.3. 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 Funds as outlined in the grant agreement? In summary, this is to confirm that the funds were used in alignment with the WMF mission and for charitable/nonprofit/educational purposes.
Yes
15. If you have additional recommendations or reflections that don’t fit into the above sections, please write them here. (optional)
N/A