{"id":9878,"date":"2022-11-05T19:28:01","date_gmt":"2022-11-05T18:28:01","guid":{"rendered":"https:\/\/www.lenseup.com\/?page_id=9878"},"modified":"2023-10-21T10:53:11","modified_gmt":"2023-10-21T08:53:11","slug":"natural-language-processing-technologies","status":"publish","type":"page","link":"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/","title":{"rendered":"Natural language processing technologies"},"content":{"rendered":"<div id='av_section_2'  class='avia-section main_color avia-section-default avia-no-shadow  avia-bg-style-scroll  avia-builder-el-0  el_before_av_promobox  avia-builder-el-first   container_wrap sidebar_right' style='background-color: #f2f2f2;  '  ><div class='container' ><main  role=\"main\" itemprop=\"mainContentOfPage\"  class='template-page content  av-content-small alpha units'><div class='post-entry post-entry-type-page post-entry-9878'><div class='entry-content-wrapper clearfix'>\n<div  style='padding-bottom:0px; color:#ffffff;font-size:40px;' class='av-special-heading av-special-heading-h1 custom-color-heading blockquote modern-quote  avia-builder-el-1  avia-builder-el-no-sibling  av-inherit-size '><h1 class='av-special-heading-tag '  itemprop=\"headline\"  >Natural language processing technologies &#8211; NLP<\/h1><div class='special-heading-border'><div class='special-heading-inner-border' style='border-color:#ffffff'><\/div><\/div><\/div>\n<\/div><\/div><\/main><!-- close content main element --><\/div><\/div><div id='after_section_1'  class='main_color av_default_container_wrap container_wrap sidebar_right' style=' '  ><div class='container' ><div class='template-page content  av-content-small alpha units'><div class='post-entry post-entry-type-page post-entry-9878'><div class='entry-content-wrapper clearfix'>\n\t<div   class='av_promobox  avia-button-no   avia-builder-el-2  el_after_av_section  el_before_av_one_half  avia-builder-el-first '>\t\t<div class='avia-promocontent'><\/p>\n<blockquote>\n<p>LenseUp uses natural language processing technologies to help solve many language issues , from content generation to semantic search or automatic speech recognition. Those are all things that we can do at LenseUp, thanks to technologies that are powered by large neural networks and state-of-the-art language AI.<\/p>\n<\/blockquote>\n<p>\n<\/div><\/div>\n<div class=\"flex_column av_one_half  flex_column_div av-zero-column-padding first  avia-builder-el-3  el_after_av_promobox  el_before_av_one_half  \" style='border-radius:0px; '><section class=\"av_textblock_section \"  itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/CreativeWork\" ><div class='avia_textblock  '   itemprop=\"text\" ><h2>NLP: how\u00a0 to analyze, understand, and generate human language with computers<\/h2>\n<p><span data-offset-key=\"2h6di-81-0\">N<\/span><span data-offset-key=\"2h6di-82-0\">LP<\/span><span data-offset-key=\"2h6di-83-0\"> is<\/span><span data-offset-key=\"2h6di-84-0\"> no<\/span><span data-offset-key=\"2h6di-85-0\"> longer<\/span><span data-offset-key=\"2h6di-86-0\"> an<\/span><span data-offset-key=\"2h6di-87-0\"> unfamiliar<\/span><span data-offset-key=\"2h6di-88-0\"> concept<\/span><span data-offset-key=\"2h6di-89-0\"> for<\/span><span data-offset-key=\"2h6di-90-0\"> many<\/span><span data-offset-key=\"2h6di-91-0\"> businesses<\/span><span data-offset-key=\"2h6di-92-0\"> that<\/span><span data-offset-key=\"2h6di-93-0\"> rely<\/span><span data-offset-key=\"2h6di-94-0\"> on<\/span><span data-offset-key=\"2h6di-95-0\"> computers<\/span><span data-offset-key=\"2h6di-96-0\">,<\/span><span data-offset-key=\"2h6di-97-0\"> as<\/span><span data-offset-key=\"2h6di-98-0\"> it<\/span><span data-offset-key=\"2h6di-99-0\"> is<\/span><span data-offset-key=\"2h6di-100-0\"> helping<\/span><span data-offset-key=\"2h6di-101-0\"> humans<\/span><span data-offset-key=\"2h6di-102-0\"> communicate<\/span><span data-offset-key=\"2h6di-103-0\"> with<\/span><span data-offset-key=\"2h6di-104-0\"> machines<\/span><span data-offset-key=\"2h6di-105-0\"> and<\/span><span data-offset-key=\"2h6di-106-0\"> vice<\/span><span data-offset-key=\"2h6di-107-0\"> versa<\/span><span data-offset-key=\"2h6di-108-0\">.<\/span><span data-offset-key=\"2h6di-109-0\"> N<\/span><span data-offset-key=\"2h6di-110-0\">LP<\/span><span data-offset-key=\"2h6di-111-0\"> is<\/span><span data-offset-key=\"2h6di-112-0\"> a<\/span><span data-offset-key=\"2h6di-113-0\"> technology<\/span><span data-offset-key=\"2h6di-114-0\"> that<\/span><span data-offset-key=\"2h6di-115-0\"> combines<\/span><span data-offset-key=\"2h6di-116-0\"> lingu<\/span><span data-offset-key=\"2h6di-117-0\">istics<\/span><span data-offset-key=\"2h6di-118-0\">,<\/span><span data-offset-key=\"2h6di-119-0\"> artificial<\/span><span data-offset-key=\"2h6di-120-0\"> intelligence<\/span><span data-offset-key=\"2h6di-121-0\">,<\/span><span data-offset-key=\"2h6di-122-0\"> and<\/span><span data-offset-key=\"2h6di-123-0\"> computer<\/span><span data-offset-key=\"2h6di-124-0\"> science<\/span><span data-offset-key=\"2h6di-125-0\"> to<\/span><span data-offset-key=\"2h6di-126-0\"> process<\/span><span data-offset-key=\"2h6di-127-0\"> and<\/span><span data-offset-key=\"2h6di-128-0\"> analyze<\/span><span data-offset-key=\"2h6di-129-0\"> large<\/span><span data-offset-key=\"2h6di-130-0\"> amounts<\/span><span data-offset-key=\"2h6di-131-0\"> of<\/span><span data-offset-key=\"2h6di-132-0\"> natural<\/span><span data-offset-key=\"2h6di-133-0\"> human<\/span><span data-offset-key=\"2h6di-134-0\"> language<\/span><span data-offset-key=\"2h6di-135-0\"> in<\/span><span data-offset-key=\"2h6di-136-0\"> various<\/span><span data-offset-key=\"2h6di-137-0\"> settings<\/span><span data-offset-key=\"2h6di-138-0\">.<\/span><\/p>\n<p><span data-offset-key=\"c6br9-81-0\">People<\/span><span data-offset-key=\"c6br9-82-0\"> tried<\/span><span data-offset-key=\"c6br9-83-0\"> to<\/span><span data-offset-key=\"c6br9-84-0\"> get<\/span><span data-offset-key=\"c6br9-85-0\"> computers<\/span><span data-offset-key=\"c6br9-86-0\"> to<\/span><span data-offset-key=\"c6br9-87-0\"> understand<\/span><span data-offset-key=\"c6br9-88-0\"> language<\/span><span data-offset-key=\"c6br9-89-0\"> by<\/span><span data-offset-key=\"c6br9-90-0\"> building<\/span><span data-offset-key=\"c6br9-91-0\"> in<\/span><span data-offset-key=\"c6br9-92-0\"> rules<\/span><span data-offset-key=\"c6br9-93-0\"> about<\/span><span data-offset-key=\"c6br9-94-0\"> the<\/span><span data-offset-key=\"c6br9-95-0\"> way<\/span><span data-offset-key=\"c6br9-96-0\"> we<\/span><span data-offset-key=\"c6br9-97-0\"> thought<\/span><span data-offset-key=\"c6br9-98-0\"> language<\/span><span data-offset-key=\"c6br9-99-0\"> worked<\/span><span data-offset-key=\"c6br9-100-0\">,<\/span><span data-offset-key=\"c6br9-101-0\"> however<\/span><span data-offset-key=\"c6br9-102-0\"> we<\/span><span data-offset-key=\"c6br9-103-0\"> could<\/span><span data-offset-key=\"c6br9-104-0\"> never<\/span><span data-offset-key=\"c6br9-105-0\"> get<\/span><span data-offset-key=\"c6br9-106-0\"> a<\/span><span data-offset-key=\"c6br9-107-0\"> computer<\/span><span data-offset-key=\"c6br9-108-0\"> to<\/span><span data-offset-key=\"c6br9-109-0\"> respond<\/span><span data-offset-key=\"c6br9-110-0\"> sens<\/span><span data-offset-key=\"c6br9-111-0\">ibly<\/span><span data-offset-key=\"c6br9-112-0\"> or<\/span><span data-offset-key=\"c6br9-113-0\"> take<\/span><span data-offset-key=\"c6br9-114-0\"> action<\/span><span data-offset-key=\"c6br9-115-0\"> based<\/span><span data-offset-key=\"c6br9-116-0\"> on<\/span><span data-offset-key=\"c6br9-117-0\"> the<\/span><span data-offset-key=\"c6br9-118-0\"> sentence<\/span><span data-offset-key=\"c6br9-119-0\"> that<\/span><span data-offset-key=\"c6br9-120-0\"> was<\/span><span data-offset-key=\"c6br9-121-0\"> given<\/span><span data-offset-key=\"c6br9-122-0\"> to<\/span><span data-offset-key=\"c6br9-123-0\"> it<\/span><span data-offset-key=\"c6br9-124-0\">.<\/span><span data-offset-key=\"c6br9-125-0\"> The<\/span><span data-offset-key=\"c6br9-126-0\"> reason<\/span><span data-offset-key=\"c6br9-127-0\"> for<\/span><span data-offset-key=\"c6br9-128-0\"> that<\/span><span data-offset-key=\"c6br9-129-0\"> is<\/span><span data-offset-key=\"c6br9-130-0\"> that<\/span><span data-offset-key=\"c6br9-131-0\"> language<\/span><span data-offset-key=\"c6br9-132-0\"> is<\/span><span data-offset-key=\"c6br9-133-0\"> very<\/span><span data-offset-key=\"c6br9-134-0\"> complex.<\/span><\/p>\n<p><span data-offset-key=\"2gcin-96-0\">However<\/span><span data-offset-key=\"2gcin-97-0\">,<\/span><span data-offset-key=\"2gcin-98-0\"> in<\/span><span data-offset-key=\"2gcin-99-0\"> the<\/span><span data-offset-key=\"2gcin-100-0\"> past<\/span><span data-offset-key=\"2gcin-101-0\"> few<\/span><span data-offset-key=\"2gcin-102-0\"> years<\/span><span data-offset-key=\"2gcin-103-0\">,<\/span><span data-offset-key=\"2gcin-104-0\"> there<\/span><span data-offset-key=\"2gcin-105-0\"> has<\/span><span data-offset-key=\"2gcin-106-0\"> been<\/span><span data-offset-key=\"2gcin-107-0\"> an<\/span><span data-offset-key=\"2gcin-108-0\"> alternative<\/span><span data-offset-key=\"2gcin-109-0\"> that<\/span><span data-offset-key=\"2gcin-110-0\"> has<\/span><span data-offset-key=\"2gcin-111-0\"> shown<\/span><span data-offset-key=\"2gcin-112-0\"> promise<\/span><span data-offset-key=\"2gcin-113-0\">:<\/span><span data-offset-key=\"2gcin-114-0\"> machine<\/span><span data-offset-key=\"2gcin-115-0\"> learning<\/span><span data-offset-key=\"2gcin-116-0\">.<\/span><span data-offset-key=\"2gcin-117-0\"> I<\/span><span data-offset-key=\"2gcin-121-0\">nstead<\/span><span data-offset-key=\"2gcin-122-0\"> of<\/span><span data-offset-key=\"2gcin-123-0\"> hard<\/span><span data-offset-key=\"2gcin-124-0\">&#8211;<\/span><span data-offset-key=\"2gcin-125-0\">c<\/span><span data-offset-key=\"2gcin-126-0\">oding<\/span><span data-offset-key=\"2gcin-127-0\"> rules<\/span><span data-offset-key=\"2gcin-128-0\">, <\/span><span data-offset-key=\"2gcin-132-0\">a<\/span><span data-offset-key=\"2gcin-133-0\"> system<\/span><span data-offset-key=\"2gcin-134-0\"> is set up to <\/span><span data-offset-key=\"2gcin-135-0\">you<\/span><span data-offset-key=\"2gcin-136-0\"> give<\/span><span data-offset-key=\"2gcin-137-0\"> the<\/span><span data-offset-key=\"2gcin-138-0\"> computer<\/span> <span data-offset-key=\"2gcin-140-0\">\u00a0many<\/span><span data-offset-key=\"2gcin-142-0\">\u00a0examples<\/span><span data-offset-key=\"2gcin-143-0\"> of<\/span><span data-offset-key=\"2gcin-144-0\"> what<\/span><span data-offset-key=\"2gcin-145-0\"> it<\/span><span data-offset-key=\"2gcin-146-0\"> should<\/span><span data-offset-key=\"2gcin-147-0\"> do<\/span><span data-offset-key=\"2gcin-148-0\">.<\/span><span data-offset-key=\"2gcin-149-0\"> The<\/span><span data-offset-key=\"2gcin-150-0\"> computer<\/span><span data-offset-key=\"2gcin-151-0\"> then<\/span><span data-offset-key=\"2gcin-152-0\"> learns<\/span><span data-offset-key=\"2gcin-153-0\"> how<\/span><span data-offset-key=\"2gcin-154-0\"> to<\/span><span data-offset-key=\"2gcin-155-0\"> do<\/span><span data-offset-key=\"2gcin-156-0\"> the<\/span><span data-offset-key=\"2gcin-157-0\"> task<\/span><span data-offset-key=\"2gcin-158-0\">.<\/span><\/p>\n<p><span data-offset-key=\"4vo7-96-0\">Ne<\/span><span data-offset-key=\"4vo7-97-0\">ural<\/span><span data-offset-key=\"4vo7-98-0\"> networks<\/span><span data-offset-key=\"4vo7-99-0\">,<\/span><span data-offset-key=\"4vo7-100-0\"> which<\/span><span data-offset-key=\"4vo7-101-0\"> are<\/span><span data-offset-key=\"4vo7-102-0\"> a<\/span><span data-offset-key=\"4vo7-103-0\"> type<\/span><span data-offset-key=\"4vo7-104-0\"> of<\/span><span data-offset-key=\"4vo7-105-0\"> machine<\/span><span data-offset-key=\"4vo7-106-0\"> learning<\/span><span data-offset-key=\"4vo7-107-0\">,<\/span><span data-offset-key=\"4vo7-108-0\"> and transformers (type of neural network) work well since they have a mechanism for looking at sequences. This is great for text. They are also easy to scale, and it is easy to make large versions, which results in good performance.<\/span><\/p>\n<\/div><\/section><\/div>\n<div class=\"flex_column av_one_half  flex_column_div av-zero-column-padding   avia-builder-el-5  el_after_av_one_half  el_before_av_layout_row  avia-builder-el-last  \" style='border-radius:0px; '><section class=\"av_textblock_section \"  itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/CreativeWork\" ><div class='avia_textblock  '   itemprop=\"text\" ><p>Data scientists now train systems by feeding them with\u00a0as much text as possible, rather than writing rules. They provide systems with texts and attempt to get them\u00a0to predict the next texst. As a result, we have a system that can take in the first half of a sentence and write the second half. It can also take in text and write summaries. Even translation can be done this way, because these are all language issues that can be handled by a system that understands the text, thanks to NLP.<\/p>\n<h2>At LenseUp, we focus on:<\/h2>\n<ul>\n<li><strong>Text generation: <\/strong>t<span data-offset-key=\"esq9e-55-0\">he<\/span><span data-offset-key=\"esq9e-56-0\"> system<\/span><span data-offset-key=\"esq9e-57-0\"> takes<\/span><span data-offset-key=\"esq9e-58-0\"> in<\/span><span data-offset-key=\"esq9e-59-0\"> text<\/span><span data-offset-key=\"esq9e-60-0\"> and<\/span><span data-offset-key=\"esq9e-61-0\"> then<\/span><span data-offset-key=\"esq9e-62-0\"> produces<\/span><span data-offset-key=\"esq9e-63-0\"> more<\/span><span data-offset-key=\"esq9e-64-0\"> text<\/span><span data-offset-key=\"esq9e-65-0\">.<\/span><span data-offset-key=\"esq9e-66-0\"> This<\/span><span data-offset-key=\"esq9e-67-0\"> approach<\/span><span data-offset-key=\"esq9e-68-0\"> can<\/span><span data-offset-key=\"esq9e-69-0\"> be<\/span><span data-offset-key=\"esq9e-70-0\"> used<\/span><span data-offset-key=\"esq9e-71-0\"> for<\/span><span data-offset-key=\"esq9e-72-0\"> many<\/span><span data-offset-key=\"esq9e-73-0\"> activities<\/span><span data-offset-key=\"esq9e-74-0\">,<\/span><span data-offset-key=\"esq9e-75-0\"> such<\/span><span data-offset-key=\"esq9e-76-0\"> as<\/span><span data-offset-key=\"esq9e-77-0\"> getting<\/span><span data-offset-key=\"esq9e-78-0\"> a<\/span><strong> summary, translation, blog, or extracting entities<\/strong><span data-offset-key=\"esq9e-88-0\">.<\/span><\/li>\n<li><strong>Embeddings:<\/strong> <span data-offset-key=\"8196q-77-0\">An<\/span><span data-offset-key=\"8196q-78-0\"> embed<\/span><span data-offset-key=\"8196q-79-0\">ding<\/span><span data-offset-key=\"8196q-80-0\"> can<\/span><span data-offset-key=\"8196q-81-0\"> be<\/span><span data-offset-key=\"8196q-82-0\"> thought<\/span><span data-offset-key=\"8196q-83-0\"> of<\/span><span data-offset-key=\"8196q-84-0\"> as<\/span><span data-offset-key=\"8196q-85-0\"> a<\/span><span data-offset-key=\"8196q-86-0\"> vector<\/span><span data-offset-key=\"8196q-87-0\">,<\/span><span data-offset-key=\"8196q-88-0\"> or<\/span><span data-offset-key=\"8196q-89-0\"> a<\/span><span data-offset-key=\"8196q-90-0\"> list<\/span><span data-offset-key=\"8196q-91-0\"> of<\/span><span data-offset-key=\"8196q-92-0\"> numbers<\/span><span data-offset-key=\"8196q-93-0\">.<\/span><span data-offset-key=\"8196q-94-0\"> So<\/span><span data-offset-key=\"8196q-95-0\">,<\/span><span data-offset-key=\"8196q-96-0\"> when<\/span><span data-offset-key=\"8196q-97-0\"> you<\/span><span data-offset-key=\"8196q-98-0\"> give<\/span><span data-offset-key=\"8196q-99-0\"> it<\/span><span data-offset-key=\"8196q-100-0\"> some<\/span><span data-offset-key=\"8196q-101-0\"> text<\/span><span data-offset-key=\"8196q-102-0\">,<\/span><span data-offset-key=\"8196q-103-0\"> it<\/span><span data-offset-key=\"8196q-104-0\"> outputs<\/span><span data-offset-key=\"8196q-105-0\"> a<\/span><span data-offset-key=\"8196q-106-0\"> list<\/span><span data-offset-key=\"8196q-107-0\"> of<\/span><span data-offset-key=\"8196q-108-0\"> numbers<\/span><span data-offset-key=\"8196q-109-0\"> that<\/span><span data-offset-key=\"8196q-110-0\"> can<\/span><span data-offset-key=\"8196q-111-0\"> be<\/span><span data-offset-key=\"8196q-112-0\"> used<\/span><span data-offset-key=\"8196q-113-0\"> for<\/span><span data-offset-key=\"8196q-114-0\"> things<\/span><span data-offset-key=\"8196q-115-0\"> such<\/span><span data-offset-key=\"8196q-116-0\"> as<\/span><span data-offset-key=\"8196q-117-0\"> semantic<\/span><span data-offset-key=\"8196q-118-0\"> search<\/span><span data-offset-key=\"8196q-119-0\"> or<\/span><span data-offset-key=\"8196q-120-0\"> clust<\/span><span data-offset-key=\"8196q-121-0\">ering<\/span><span data-offset-key=\"8196q-122-0\">.<\/span><span data-offset-key=\"8196q-123-0\"> This<\/span><span data-offset-key=\"8196q-124-0\"> is<\/span><span data-offset-key=\"8196q-125-0\"> done<\/span><span data-offset-key=\"8196q-126-0\"> by<\/span><span data-offset-key=\"8196q-127-0\"> measuring<\/span><span data-offset-key=\"8196q-128-0\"> the<\/span><span data-offset-key=\"8196q-129-0\"> distance<\/span><span data-offset-key=\"8196q-130-0\"> within<\/span><span data-offset-key=\"8196q-131-0\"> the<\/span><span data-offset-key=\"8196q-132-0\"> vector<\/span><span data-offset-key=\"8196q-133-0\"> space<\/span><span data-offset-key=\"8196q-134-0\">.<\/span> This approach is very useful and has many applications such as<strong> semantic search for Question \/ Answer chatbots.<\/strong><\/li>\n<li><strong>Multilingual NLP:<\/strong> one<span data-offset-key=\"93gmo-143-0\"> of<\/span><span data-offset-key=\"93gmo-144-0\"> the<\/span><span data-offset-key=\"93gmo-145-0\"> main<\/span><span data-offset-key=\"93gmo-146-0\"> reasons<\/span><span data-offset-key=\"93gmo-147-0\"> why<\/span><span data-offset-key=\"93gmo-148-0\"> mult<\/span><span data-offset-key=\"93gmo-149-0\">ilingual<\/span><span data-offset-key=\"93gmo-150-0\"> N<\/span><span data-offset-key=\"93gmo-151-0\">LP<\/span><span data-offset-key=\"93gmo-152-0\"> has<\/span><span data-offset-key=\"93gmo-153-0\"> not<\/span><span data-offset-key=\"93gmo-154-0\"> been<\/span><span data-offset-key=\"93gmo-155-0\"> able<\/span><span data-offset-key=\"93gmo-156-0\"> to<\/span><span data-offset-key=\"93gmo-157-0\"> scale<\/span><span data-offset-key=\"93gmo-158-0\"> quickly<\/span><span data-offset-key=\"93gmo-159-0\"> is<\/span><span data-offset-key=\"93gmo-160-0\"> due<\/span><span data-offset-key=\"93gmo-161-0\"> to<\/span><span data-offset-key=\"93gmo-162-0\"> the<\/span><span data-offset-key=\"93gmo-163-0\"> lack<\/span><span data-offset-key=\"93gmo-164-0\"> of<\/span><span data-offset-key=\"93gmo-165-0\"> labelled<\/span><span data-offset-key=\"93gmo-166-0\"> data<\/span><span data-offset-key=\"93gmo-167-0\"> in<\/span><span data-offset-key=\"93gmo-168-0\"> low<\/span><span data-offset-key=\"93gmo-169-0\">&#8211;<\/span><span data-offset-key=\"93gmo-170-0\">resource<\/span><span data-offset-key=\"93gmo-171-0\"> languages<\/span><span data-offset-key=\"93gmo-172-0\">.<\/span><span data-offset-key=\"93gmo-173-0\"> BL<\/span><span data-offset-key=\"93gmo-174-0\">O<\/span><span data-offset-key=\"93gmo-175-0\">OM<\/span><span data-offset-key=\"93gmo-176-0\"> is<\/span><span data-offset-key=\"93gmo-177-0\"> the<\/span><span data-offset-key=\"93gmo-178-0\"> largest<\/span><span data-offset-key=\"93gmo-179-0\"> mult<\/span><span data-offset-key=\"93gmo-180-0\">ilingual<\/span><span data-offset-key=\"93gmo-181-0\"> language<\/span><span data-offset-key=\"93gmo-182-0\"> model<\/span><span data-offset-key=\"93gmo-183-0\"> to<\/span><span data-offset-key=\"93gmo-184-0\"> be<\/span><span data-offset-key=\"93gmo-185-0\"> trained<\/span><span data-offset-key=\"93gmo-186-0\"> openly<\/span><span data-offset-key=\"93gmo-187-0\"> and<\/span><span data-offset-key=\"93gmo-188-0\"> transparent<\/span><span data-offset-key=\"93gmo-189-0\">ly<\/span><span data-offset-key=\"93gmo-190-0\">,<\/span><span data-offset-key=\"93gmo-191-0\"> which<\/span><span data-offset-key=\"93gmo-192-0\"> may<\/span><span data-offset-key=\"93gmo-193-0\"> help<\/span><span data-offset-key=\"93gmo-194-0\"> to<\/span><span data-offset-key=\"93gmo-195-0\"> solve<\/span><span data-offset-key=\"93gmo-196-0\"> this<\/span><span data-offset-key=\"93gmo-197-0\"> issue<\/span><span data-offset-key=\"93gmo-198-0\">. It was released in july 2022. Multilingual NLP, such as we can witness it in models such as Openai&#8217;s Whisper, is changing the game!<\/span><\/li>\n<\/ul>\n<\/div><\/section><\/div>\n<\/div><\/div><\/div><!-- close content main div --><\/div><\/div><div id='best-rated'  class='av-layout-grid-container entry-content-wrapper main_color av-flex-cells     avia-builder-el-7  el_after_av_one_half  el_before_av_section  submenu-not-first container_wrap sidebar_right' style=' '  >\n<div class=\"flex_cell no_margin av_one_half  avia-builder-el-8  el_before_av_cell_one_half  avia-builder-el-first   \"  style='vertical-align:middle; padding:50px 80px 50px 6% ; background-color:#f8f8f8; ' ><div class='flex_cell_inner' ><section class=\"av_textblock_section \"  itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/CreativeWork\" ><div class='avia_textblock  '   itemprop=\"text\" ><h2>For which uses?<\/h2>\n<\/div><\/section><br \/>\n<section class=\"av_textblock_section \"  itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/CreativeWork\" ><div class='avia_textblock  '   itemprop=\"text\" ><p>The overarching goal in all these cases is to take raw language input and use linguistics and algorithms to transform or enrich the text so that it delivers greater value.<\/p>\n<ul>\n<li><strong>Topic discovery<\/strong><\/li>\n<li><strong>Sentiment analysis<\/strong><\/li>\n<li><strong>Machine translation<\/strong><\/li>\n<li><strong>Audio summarization<\/strong><\/li>\n<li><strong>Speech-to-text <\/strong><\/li>\n<li><strong>Text-to-speech\u00a0<\/strong><\/li>\n<li><strong>Conversational a.i. and chatbots<\/strong><\/li>\n<\/ul>\n<\/div><\/section><br \/>\n<section class=\"av_textblock_section \"  itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/CreativeWork\" ><div class='avia_textblock  '   itemprop=\"text\" ><h2><span class=\"section-heading\">Connect with LenseUp and see what we can do for you.<\/span><\/h2>\n<p><span data-offset-key=\"9ipnv-111-0\">The<\/span><span data-offset-key=\"9ipnv-112-0\"> complexities<\/span><span data-offset-key=\"9ipnv-113-0\"> and<\/span><span data-offset-key=\"9ipnv-114-0\"> diversity<\/span><span data-offset-key=\"9ipnv-115-0\"> of<\/span><span data-offset-key=\"9ipnv-116-0\"> human<\/span><span data-offset-key=\"9ipnv-117-0\"> language<\/span><span data-offset-key=\"9ipnv-118-0\"> is<\/span><span data-offset-key=\"9ipnv-119-0\"> astounding<\/span><span data-offset-key=\"9ipnv-120-0\">.<\/span><span data-offset-key=\"9ipnv-121-0\"> We<\/span><span data-offset-key=\"9ipnv-122-0\"> express<\/span><span data-offset-key=\"9ipnv-123-0\"> ourselves<\/span><span data-offset-key=\"9ipnv-124-0\"> in<\/span><span data-offset-key=\"9ipnv-125-0\"> an<\/span><span data-offset-key=\"9ipnv-126-0\"> infinite<\/span><span data-offset-key=\"9ipnv-127-0\"> number<\/span><span data-offset-key=\"9ipnv-128-0\"> of<\/span><span data-offset-key=\"9ipnv-129-0\"> ways<\/span><span data-offset-key=\"9ipnv-130-0\">,<\/span><span data-offset-key=\"9ipnv-131-0\"> both<\/span><span data-offset-key=\"9ipnv-132-0\"> verbally<\/span><span data-offset-key=\"9ipnv-133-0\"> and<\/span><span data-offset-key=\"9ipnv-134-0\"> in<\/span><span data-offset-key=\"9ipnv-135-0\"> writing<\/span><span data-offset-key=\"9ipnv-136-0\">.<\/span><span data-offset-key=\"9ipnv-137-0\"> Not<\/span><span data-offset-key=\"9ipnv-138-0\"> only<\/span><span data-offset-key=\"9ipnv-139-0\"> are<\/span><span data-offset-key=\"9ipnv-140-0\"> there<\/span><span data-offset-key=\"9ipnv-141-0\"> hundreds<\/span><span data-offset-key=\"9ipnv-142-0\"> of<\/span><span data-offset-key=\"9ipnv-143-0\"> languages<\/span><span data-offset-key=\"9ipnv-144-0\"> and<\/span><span data-offset-key=\"9ipnv-145-0\"> dialect<\/span><span data-offset-key=\"9ipnv-146-0\">s<\/span><span data-offset-key=\"9ipnv-147-0\">,<\/span><span data-offset-key=\"9ipnv-148-0\"> but<\/span><span data-offset-key=\"9ipnv-149-0\"> within<\/span><span data-offset-key=\"9ipnv-150-0\"> each<\/span><span data-offset-key=\"9ipnv-151-0\"> language<\/span><span data-offset-key=\"9ipnv-152-0\"> is<\/span><span data-offset-key=\"9ipnv-153-0\"> a<\/span><span data-offset-key=\"9ipnv-154-0\"> unique<\/span><span data-offset-key=\"9ipnv-155-0\"> set<\/span><span data-offset-key=\"9ipnv-156-0\"> of<\/span><span data-offset-key=\"9ipnv-157-0\"> grammar<\/span><span data-offset-key=\"9ipnv-158-0\"> and<\/span><span data-offset-key=\"9ipnv-159-0\"> syntax<\/span><span data-offset-key=\"9ipnv-160-0\"> rules<\/span><span data-offset-key=\"9ipnv-161-0\">,<\/span><span data-offset-key=\"9ipnv-162-0\"> terms<\/span><span data-offset-key=\"9ipnv-163-0\"> and<\/span><span data-offset-key=\"9ipnv-164-0\"> slang<\/span><span data-offset-key=\"9ipnv-165-0\">.<\/span><span data-offset-key=\"9ipnv-166-0\"> When<\/span><span data-offset-key=\"9ipnv-167-0\"> we<\/span><span data-offset-key=\"9ipnv-168-0\"> write<\/span><span data-offset-key=\"9ipnv-169-0\">,<\/span><span data-offset-key=\"9ipnv-170-0\"> we<\/span><span data-offset-key=\"9ipnv-171-0\"> often<\/span><span data-offset-key=\"9ipnv-172-0\"> miss<\/span><span data-offset-key=\"9ipnv-173-0\">pell<\/span><span data-offset-key=\"9ipnv-174-0\"> or<\/span><span data-offset-key=\"9ipnv-175-0\"> abbrevi<\/span><span data-offset-key=\"9ipnv-176-0\">ate<\/span><span data-offset-key=\"9ipnv-177-0\"> words<\/span><span data-offset-key=\"9ipnv-178-0\">,<\/span><span data-offset-key=\"9ipnv-179-0\"> or<\/span><span data-offset-key=\"9ipnv-180-0\"> omit<\/span><span data-offset-key=\"9ipnv-181-0\"> punct<\/span><span data-offset-key=\"9ipnv-182-0\">uation<\/span><span data-offset-key=\"9ipnv-183-0\">.<\/span><span data-offset-key=\"9ipnv-184-0\"> When<\/span><span data-offset-key=\"9ipnv-185-0\"> we<\/span><span data-offset-key=\"9ipnv-186-0\"> speak<\/span><span data-offset-key=\"9ipnv-187-0\">,<\/span><span data-offset-key=\"9ipnv-188-0\"> we<\/span><span data-offset-key=\"9ipnv-189-0\"> have<\/span><span data-offset-key=\"9ipnv-190-0\"> regional<\/span><span data-offset-key=\"9ipnv-191-0\"> accents<\/span><span data-offset-key=\"9ipnv-192-0\">,<\/span><span data-offset-key=\"9ipnv-193-0\"> and<\/span><span data-offset-key=\"9ipnv-194-0\"> we<\/span><span data-offset-key=\"9ipnv-195-0\"> m<\/span><span data-offset-key=\"9ipnv-196-0\">umble<\/span><span data-offset-key=\"9ipnv-197-0\">,<\/span><span data-offset-key=\"9ipnv-198-0\"> st<\/span><span data-offset-key=\"9ipnv-199-0\">utter<\/span><span data-offset-key=\"9ipnv-200-0\"> and<\/span><span data-offset-key=\"9ipnv-201-0\"> borrow<\/span><span data-offset-key=\"9ipnv-202-0\"> terms<\/span><span data-offset-key=\"9ipnv-203-0\"> from<\/span><span data-offset-key=\"9ipnv-204-0\"> other<\/span><span data-offset-key=\"9ipnv-205-0\"> languages<\/span><span data-offset-key=\"9ipnv-206-0\">.<\/span><\/p>\n<p><span data-offset-key=\"45dqc-50-0\">N<\/span><span data-offset-key=\"45dqc-51-0\">LP<\/span><span data-offset-key=\"45dqc-52-0\"> is<\/span><span data-offset-key=\"45dqc-53-0\"> beneficial<\/span><span data-offset-key=\"45dqc-54-0\"> because<\/span><span data-offset-key=\"45dqc-55-0\"> it<\/span><span data-offset-key=\"45dqc-56-0\"> eliminated<\/span><span data-offset-key=\"45dqc-57-0\"> ambiguity<\/span><span data-offset-key=\"45dqc-58-0\"> in<\/span><span data-offset-key=\"45dqc-59-0\"> language<\/span><span data-offset-key=\"45dqc-60-0\"> and<\/span><span data-offset-key=\"45dqc-61-0\"> creates<\/span><span data-offset-key=\"45dqc-62-0\"> necessary<\/span><span data-offset-key=\"45dqc-63-0\"> numeric<\/span><span data-offset-key=\"45dqc-64-0\"> structure<\/span><span data-offset-key=\"45dqc-65-0\"> for<\/span><span data-offset-key=\"45dqc-66-0\"> data<\/span><span data-offset-key=\"45dqc-67-0\"> in<\/span><span data-offset-key=\"45dqc-68-0\"> many<\/span><span data-offset-key=\"45dqc-69-0\"> applications<\/span><span data-offset-key=\"45dqc-70-0\">,<\/span><span data-offset-key=\"45dqc-71-0\"> such<\/span><span data-offset-key=\"45dqc-72-0\"> as<\/span><span data-offset-key=\"45dqc-73-0\"> speech<\/span><span data-offset-key=\"45dqc-74-0\"> recognition<\/span><span data-offset-key=\"45dqc-75-0\"> or<\/span><span data-offset-key=\"45dqc-76-0\"> text<\/span><span data-offset-key=\"45dqc-77-0\"> summarization.<\/span><\/p>\n<\/div><\/section><br \/>\n<div  style='height:20px' class='hr hr-invisible   avia-builder-el-12  el_after_av_textblock  el_before_av_hr '><span class='hr-inner ' ><span class='hr-inner-style'><\/span><\/span><\/div><br \/>\n<div  style='height:42px' class='hr hr-invisible   avia-builder-el-13  el_after_av_hr  el_before_av_button '><span class='hr-inner ' ><span class='hr-inner-style'><\/span><\/span><\/div><br \/>\n<div  class='avia-button-wrap avia-button-left  avia-builder-el-14  el_after_av_hr  avia-builder-el-last ' ><a href='#section2'  class='avia-button av-icon-on-hover avia-color-theme-color  avia-font-color-theme-color  avia-icon_select-yes-left-icon avia-size-large avia-position-left '   ><span class='avia_button_icon avia_button_icon_left ' aria-hidden='true' data-av_icon='\ue859' data-av_iconfont='entypo-fontello'><\/span><span class='avia_iconbox_title' >Request a quote<\/span><span class='avia_button_background avia-button avia-color-theme-color-highlight' ><\/span><\/a><\/div><\/p>\n<\/div><\/div><div class=\"flex_cell no_margin av_one_half  avia-builder-el-15  el_after_av_cell_one_half  avia-builder-el-last   avia-full-stretch \"  style='background:url(https:\/\/www.lenseup.com\/wp-content\/uploads\/2022\/02\/neural-machine-translation.jpg) bottom left no-repeat scroll #f8f8f8; vertical-align:middle; padding:200px 0px 200px 0px ; background-color:#f8f8f8; ' ><div class='flex_cell_inner' ><\/div><\/div>\n<\/div>\n<div id='my_section'  class='avia-section main_color avia-section-no-padding avia-no-shadow  avia-bg-style-scroll  avia-builder-el-16  el_after_av_layout_row  el_before_av_textblock   container_wrap sidebar_right' style='background-color: #e72b2b;  '  ><div class='container' ><div class='template-page content  av-content-small alpha units'><div class='post-entry post-entry-type-page post-entry-9878'><div class='entry-content-wrapper clearfix'><\/div><\/div><\/div><!-- close content main div --><\/div><\/div><div id='after_section_2'  class='main_color av_default_container_wrap container_wrap sidebar_right' style=' '  ><div class='container' ><div class='template-page content  av-content-small alpha units'><div class='post-entry post-entry-type-page post-entry-9878'><div class='entry-content-wrapper clearfix'>\n<section class=\"av_textblock_section \"  itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/CreativeWork\" ><div class='avia_textblock  '   itemprop=\"text\" ><h2 id=\"section2\">Ask for a quote<\/h2>\n<\/div><\/section>\n\n<\/div><\/div><\/div><!-- close content main div --><\/div><\/div><div id='my_section'  class='avia-section main_color avia-section-no-padding avia-no-shadow  avia-bg-style-scroll  avia-builder-el-19  el_after_av_codeblock  avia-builder-el-last   container_wrap sidebar_right' style='background-color: #e72b2b;  '  ><div class='container' ><div class='template-page content  av-content-small alpha units'><div class='post-entry post-entry-type-page post-entry-9878'><div class='entry-content-wrapper clearfix'><\/div><\/div><\/div><!-- close content main div --><\/div><\/div><div id='after_section_3'  class='main_color av_default_container_wrap container_wrap sidebar_right' style=' '  ><div class='container' ><div class='template-page content  av-content-small alpha units'><div class='post-entry post-entry-type-page post-entry-9878'><div class='entry-content-wrapper clearfix'>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v23.5 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>NLP &amp; Machine Learning Agency in France<\/title>\n<meta name=\"description\" content=\"LenseUp uses natural language processing technologies to help solve many language issues , from content generation to semantic search, or speech recognition.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"NLP &amp; Machine Learning Agency in France\" \/>\n<meta property=\"og:description\" content=\"LenseUp uses natural language processing technologies to help solve many language issues , from content generation to semantic search, or speech recognition.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/\" \/>\n<meta property=\"og:site_name\" content=\"LenseUp, video and audio solutions\" \/>\n<meta property=\"article:modified_time\" content=\"2023-10-21T08:53:11+00:00\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"9 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/\",\"url\":\"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/\",\"name\":\"NLP & Machine Learning Agency in France\",\"isPartOf\":{\"@id\":\"https:\/\/www.lenseup.com\/en\/#website\"},\"datePublished\":\"2022-11-05T18:28:01+00:00\",\"dateModified\":\"2023-10-21T08:53:11+00:00\",\"description\":\"LenseUp uses natural language processing technologies to help solve many language issues , from content generation to semantic search, or speech recognition.\",\"breadcrumb\":{\"@id\":\"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Accueil\",\"item\":\"https:\/\/www.lenseup.com\/en\/home-oct-2021\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Natural language processing technologies\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.lenseup.com\/en\/#website\",\"url\":\"https:\/\/www.lenseup.com\/en\/\",\"name\":\"LenseUp, multilingual audio and video solutions\",\"description\":\"Audioguides, audio books, audio and video translations, multilingual chatbots... discover LenseUp.\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.lenseup.com\/en\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"NLP & Machine Learning Agency in France","description":"LenseUp uses natural language processing technologies to help solve many language issues , from content generation to semantic search, or speech recognition.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/","og_locale":"en_US","og_type":"article","og_title":"NLP & Machine Learning Agency in France","og_description":"LenseUp uses natural language processing technologies to help solve many language issues , from content generation to semantic search, or speech recognition.","og_url":"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/","og_site_name":"LenseUp, video and audio solutions","article_modified_time":"2023-10-21T08:53:11+00:00","twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"9 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/","url":"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/","name":"NLP & Machine Learning Agency in France","isPartOf":{"@id":"https:\/\/www.lenseup.com\/en\/#website"},"datePublished":"2022-11-05T18:28:01+00:00","dateModified":"2023-10-21T08:53:11+00:00","description":"LenseUp uses natural language processing technologies to help solve many language issues , from content generation to semantic search, or speech recognition.","breadcrumb":{"@id":"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.lenseup.com\/en\/natural-language-processing-technologies\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Accueil","item":"https:\/\/www.lenseup.com\/en\/home-oct-2021\/"},{"@type":"ListItem","position":2,"name":"Natural language processing technologies"}]},{"@type":"WebSite","@id":"https:\/\/www.lenseup.com\/en\/#website","url":"https:\/\/www.lenseup.com\/en\/","name":"LenseUp, multilingual audio and video solutions","description":"Audioguides, audio books, audio and video translations, multilingual chatbots... discover LenseUp.","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.lenseup.com\/en\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"}]}},"_links":{"self":[{"href":"https:\/\/www.lenseup.com\/en\/wp-json\/wp\/v2\/pages\/9878"}],"collection":[{"href":"https:\/\/www.lenseup.com\/en\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.lenseup.com\/en\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.lenseup.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.lenseup.com\/en\/wp-json\/wp\/v2\/comments?post=9878"}],"version-history":[{"count":14,"href":"https:\/\/www.lenseup.com\/en\/wp-json\/wp\/v2\/pages\/9878\/revisions"}],"predecessor-version":[{"id":14263,"href":"https:\/\/www.lenseup.com\/en\/wp-json\/wp\/v2\/pages\/9878\/revisions\/14263"}],"wp:attachment":[{"href":"https:\/\/www.lenseup.com\/en\/wp-json\/wp\/v2\/media?parent=9878"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}