Improving search accuracy and efficiency with Dicta

At Sefaria, we’re always working to improve and upgrade our technology to provide our users with the fastest and most effective methods of Torah study. That’s why today, we’re happy to announce that we are now using the Tanakh search service from Dicta to return Tanakh search results for Hebrew searches. Dicta is a non-profit organization that provides its products at no charge for the benefit of the public, and they’ve been a fellow-traveler and help for Sefaria all along the way. Now, using Dicta's technology, searches will better account for Hebrew verb conjugation, alternate spellings, prefixes and suffixes, numbers, and more. 

For example:

  • Search for “הלך” and now you’ll find לֶךְ־לְךָ֛

  • Search for 12 and you’ll find things like:

וּבִשְׁנֵים֩ עָשָׂ֨ר חֹ֜דֶשׁ הוּא־חֹ֣דֶשׁ אֲדָ֗ר בִּשְׁלֹושָׁ֨ה עָשָׂ֥ר יֹום֙ בֹּ֔ו אֲשֶׁ֨ר הִגִּ֧יעַ דְּבַר־הַמֶּ֛לֶךְ וְדָתֹ֖ו לְהֵעָשֹׂ֑ות בַּיֹּ֗ום אֲשֶׁ֨ר שִׂבְּר֜וּ אֹיְבֵ֤י הַיְּהוּדִים֙ לִשְׁלֹ֣וט בָּהֶ֔ם וְנַהֲפֹ֣וךְ ה֔וּא אֲשֶׁ֨ר יִשְׁלְט֧וּ הַיְּהוּדִ֛ים הֵ֖מָּה בְּשֹׂנְאֵיהֶֽם׃

  • In the past, if you searched for מוריה the first result would be II Chronicles 3:1.  We missed Genesis 22:2, since the word is written הַמֹּרִיָּה.

וַיֹּאמֶר קַח־נָא אֶת־בִּנְךָ אֶת־יְחִידְךָ אֲשֶׁר־אָהַבְתָּ אֶת־יִצְחָק וְלֶךְ־לְךָ אֶל־אֶרֶץ הַמֹּרִיָּה וְהַעֲלֵהוּ שָׁם לְעֹלָה עַל אַחַד הֶהָרִים אֲשֶׁר אֹמַר אֵלֶיךָ׃

  • If you searched for לב, we would only find verses where it was written just like this.  Now, we also find לבי, לבו, לבכם and similar instances. Like Genesis 18:5:

וְאֶקְחָה פַת־לֶחֶם וְסַעֲדוּ לִבְּכֶם אַחַר תַּעֲבֹרוּ כִּי־עַל־כֵּן עֲבַרְתֶּם עַל־עַבְדְּכֶם וַיֹּאמְרוּ כֵּן תַּעֲשֶׂה כַּאֲשֶׁר דִּבַּרְתָּ׃

  • And let’s be honest.  Sometimes our morphology was just plain weird.  Like searching for אריה. Why did we return Exodus 21:10 as the top result?

אִם־אַחֶרֶת יִקַּח־לוֹ שְׁאֵרָהּ כְּסוּתָהּ וְעֹנָתָהּ לֹא יִגְרָע׃

Now, the results make more sense, with Genesis 49:9 topping the list

גּוּר אַרְיֵה יְהוּדָה מִטֶּרֶף בְּנִי עָלִיתָ כָּרַע רָבַץ כְּאַרְיֵה וּכְלָבִיא מִי יְקִימֶנּוּ׃

We’re very thankful that Dicta has made this service available.