Nouveau corpus, nouvelles heures

Vingt registres supplémentaires ont été intégrés à l’interface de recherche. Parmi les nouvelles recherches possibles, nous trouvons une nouvelle preuve de la présence quotidienne des livres d’heures.

Paris, Arch. nat., JJ 179, fol. 81v, n° 150

A la Saint-Martin, le 11 novembre, 1447, Raoul est trouvé noyé dans un estang. Suicide ou accident ? Ses parents demandent grâce au roi pour que le doute ne jette pas le déshonneur sur la famille et n’empêche pas l’application des règles normales d’héritage. Que reste-t-il dudit Raoul au bord de l’étang ? Son livre d’heures, son manteau, ses chaussures et son chapeau (« et ses heures, mantel, patins et chappel estans sur ladicte chausee pres de la bonde dudit estang ») !
Revenant du Mexique, je souhaite à tous une bonne fête des Morts…

Concluding conference of the HIMANIS project on 27 October 2017

To celebrate the achievements and the end of the HIMANIS project, the Institut de Recherche et d’Histoire des Textes (IRHT) is organizng a conference on Friday, 27 October 2017, to present, explain and discuss the impact of handwriting recognition technologies for medieval studies and heritage institutions.

The European project HIMANIS (HIstorical MANuscript Indexing for user-controlled search), led by the IRHT, brought together several research teams in engineering sciences, both private (A2iA, France), and public (PRHLT-UPVLC, Spain and Alice-Groningen Univ., The Netherlands), with the support of the National Archives and the National Library of France.

The main result of this project is that it is now possible to carry out keyword searches in the corpus of registers of the French royal chancery of the 14th and 15th centuries. For the moment there are 167 volumes of the National Archives series, JJ 7 – JJ 209 and 20 volumes of the Bibliothèque nationale de France. The interface (still preliminary) can be accessed at the following address: http://prhlt-kws.prhlt.upv.es/himanis/. It is a test interface, which will evolve in the coming months, with additional manuscripts and new functionalities, notably to
navigate between the results, to display the regests and editions. HIMANIS also aims at computer-assisted analysis of the different hands involved in each volume, to study the functioning of the royal chancellery and the progressive construction of the centralized state of the French monarchy.

Entry is free but registration is required: Registration form.

Place

IRHT
40 Avenue d’Iéna
75016 Paris

Program

14h – François Bougard (Director of IRHT) and Emmanuel Rousseau (Director of collections, National Archives), Welcome and introduction.

14h30 – Jean-François Moufflet (Heritage Curator in the Department of the Middle Ages and the Ancien Régime, National Archives), Le projet HIMANIS ou la mise à l’index des registres.

15h – Dominique Stutzmann (researcher, IRHT), Automatic indexing for research in medieval history.

15h30 –  Christopher Kermorvant (A2iA/Teklia), Artificial paleography : training computer to read medieval manuscripts.

4 pm – Break
16:15 – Enrique Vidal (Universitat Politècnica de València), Text indexing and user-controlled search.

16h45 – Lambert Schomaker (University Groningen) and Sukalpa Chanda (University Groningen), Writer identification.

17h15-18h – Roundtable with researchers and representatives of end users, with Sébastien Barret (head of the Diplomatic Section, IRHT), Paul Bertrand (professor of medieval history and digital humanities at the Catholic University of Louvain), Charlotte Denoël (curator of libraries, head of the department of medieval manuscripts, Bibliothèque Nationale de France) and Judicael Petrowiste (lecturer in medieval history (University of Paris 7 – Paris Diderot, to be confirmed).

The Royal Highness / L’altesse royale / regalis celsitudo : une thématique de préambule pour mesurer l’utilité d’HIMANIS

Il y a quelques années, deux de mes amis ont publié une somme magistrale sur les préambules des rois de France, étudiant la production des actes royaux et ses acteurs au regard des actions juridiques et des thématiques et du style rhétorique des préambules : Barret, Sébastien, et Benoît Grévin. Regalis excellentia :  les préambules des actes des rois de France au XIVe siècle, 1300-1380. Paris: École des chartes, 2014 (Mémoires et documents de l’École des chartes, 98), ci-après abrégé « Barret-Grévin ».

La finesse de ce travail, sa complétude et son existence même, preuve que les méthodes les plus austères de l’érudition ouvrent des vues riches et chatoyantes sur les pratiques intellectuelles et l’histoire culturelle.

Pour cette importante étude, les auteurs ont notamment dépouillé les registres de la chancellerie royale produits durant le règne de Jean II le Bon, offrant au lecteur un très important corpus de préambules, comprenant  548 numéros (p. 425-576), dont trente-et-un font l’objet d’une analyse stylistique prouvant et décomposant les influences de l’ars dictaminis dans l’écriture des notaires et secrétaires du roi (p. 577-623). Les champs lexicaux et sémantiques sont illustrés et mesurés dans le corps de l’étude, mais aussi dans une annexe très fournie : aux p. 625-643, champs sémantiques ; aux p. 643-811, répertoire lexical des termes latins ; aux p. 812-828, répertoire lexical des termes français.

Pour ce billet, c’est principalement la partie documentaire de l’ouvrage qui sera utilisée, afin de mesurer l’utilité des recherches et développements faits dans le projet HIMANIS, même si je tenterai, in fine, quelques remarques sur les préambules eux-mêmes.

Méthodologie : performance, précision, rappel et résultats

Après l’énorme travail d’indexation, il faut évaluer la performance de l’indexation. Nous allons donc comparer nos résultats d’indexation automatique au dépouillement de nos collègues sur une tranche de cotes (JJ 80 – JJ 93 et JJ 105, avec JJ 81 dépouillé en partie seulement).

Méthodologie

Pour mesurer la performance de l’indexation, il faut mesurer deux valeurs distinctes : d’une part la « précision », c’est-à-dire le pourcentage d’occurrences correctes parmi l’ensemble des résultats fournis à une requête, et, d’autre part, le « rappel », c’est-à-dire le pourcentage d’occurrences correctes trouvées parmi celles que le système aurait dû trouver dans le corpus indexé.

C’est là que le dépouillement systématique des registres par Sébastien Barret et Benoît Grévin se révèle extrêmement précieux. En effet, le « rappel » est difficile à mesurer, puisqu’il faut avoir lu les registres pour fournir la « vérité terrain » et savoir combien d’occurrences se trouvent réellement dans le corpus.

La méthode suivante a été mise en œuvre. Dans un premier temps, j’ai exploité l’index et le corpus de préambules de l’ouvrage imprimé :

  • choix d’un concept dans l’index Barret-Grévin, en l’occurrence « celsitudo » qui renvoie à 45 numéros du corpus ;
  • exploitation du corpus, repérage des actes multiples utilisant un même préambule ;
  • établissement de la liste des cotes avec préambules contenant le terme recherché.

Dans un deuxième temps, j’ai utilisé l’interface de recherche http://prhlt-kws.prhlt.upv.es/himanis/ avec un « niveau de confiance » (« confidence ») placé à 0 :

  • recherche successive des termes « celsitudo » (56 occurrences), « celsitudinem » (64 occurrences), « celsitudinis » (44 occurrences), « celsitudini » (29 occurrences), « celsitudine » (38 occurrences), et validation systématique : cette étape permet de mesurer de façon fine la précision, ce qui n’est pas réellement possible si on fait une recherche combinée avec plusieurs termes ;
  • recherche avec les cinq termes avec l’opérateur ||, pour « OU », menant à 194 occurrences reconnues du mot sur les images, au lieu des 231 qui sont la somme des différentes indexations : le système lui-même en compte davantage, car il compte une occurrence de « celsitudini » indexée sous « celsitudinis » et sous « celsitudini » pour deux occurrences distinctes ;
  • lecture des textes ainsi indexés pour identifier si le mot se trouve dans un préambule ou dans une autre partie de l’acte ;
  • le cas échéant, vérification dans le corpus Barret-Grévin et identification du numéro de corps.

Enfin, dans un troisième temps, pour mesurer le « rappel », il convenait de vérifier les occurrences de préambules présents dans le corpus et non retrouvé par notre système pour essayer de comprendre le « mécanisme de la faute ».

Mesure de la précision

L’une des difficultés avec l’indexation profonde, c’est que chaque mot peut faire l’objet d’indexations multiples. Parfois, la requête « celsitudinis » aboutira à ce que l’on appelle un « faux positif », par exemple une occurrence de « celsitudini ». Mais la forme qui a été incorrectement indexée, menant au « faux positif » peut également être indexée sous l’entrée correcte. Parfois, le faux positif est la seule entrée de l’index automatique, par exemple il arrive que « celsitudinem » avec tilde final soit indexé sous « celsitudine », comme si le tilde n’avait pas été vu. Dans un cas (JJ56, f. 57r), « celsitudinem » avec tilde final est indexé sous « celsitudine » (sans tilde) et sous « celsitudinere », qui est, certes, un barbarisme, mais manifeste correctement la perception d’un tilde, mal interprété.

Le tableau suivant montre la mesure de la précision à la fois de façon stricte (« faux positif »), mais aussi dans une approche lemmatisée, c’est-à-dire en acceptant comme occurrence correcte une indexation qui n’est pas exacte, mais qui concorde avec l’une des formes déclinées de « celsitudo », par exemple si le système a indexé « sollicitudinis », alors que le texte porte « sollicitudini », on a compté ce dernier comme un « vrai positif ».

 

Entrée d’indexation Occurrences Faux positif Faux positif (hors déclinaison) Précision Précision (hors déclinaison)
celsitudo 56 0 0 1 1
celsitudinem 62 4 0 0,93548387 1
celsitudinis 44 8 0 0,81818182 1
celsitudini 29 1 0 0,96551724 1
celsitudine 38 22 0 0,42105263 1
TOTAL 229 35 0 0,84716157 1

 

La précision moyenne est de 84,7 % de façon stricte et de 100 % si l’on intègre les autres formes déclinées ! C’est d’autant plus admirable que j’avais, il faut le rappeler, placer la « confiance » à zéro, ce qui n’a pas mené à produire du bruit informationnel et la précision est plus que satisfaisante.

Rappel

Ce que je redoutais, par rapport à la précision que nous savons déjà excellente (cf. billets précédents), c’est le rappel. Pour cette mesure, il faut d’abord compter les occurrences des 45 préambules attestées dans les registres. L’un des préambules recensés dans l’index de l’ouvrage imprimé l’est sous un numéro fautif et il faut remplacer le n° 74 (indiqué dans l’index Barret-Grévin et renvoyant à JJ 89, fol. 325r, n° 674) qui ne contient pas le mot « celsitudo » par le n° 75, renvoyant à un formulaire. En tout, il y a 59 occurrences alléguées par les notices des préambules recensés dans l’index.

Comme les auteurs le mentionnent (p. 427), le renvoi à une édition sans indication de cote ne suppose pas l’absence de l’acte dans les registres eux-mêmes, mais, généralement, quand le préambule est connue par une autre source, la cote du registre n’est pas indiquée. Il y reste en tout 35 occurrences dans les registres du Trésor des chartes.

Parmi ces 35, nous en identifions 17, soit 48,6 %. Des circonstances permettent de non pas seulement de se consoler de ce résultat un peu décevant, mais de le corriger en l’améliorant.

Tout d’abord, on trouve trois occurrences de préambules que nous ne cherchions pas car cités d’après une édition ou une copie tardive : les préambules Barret-Grévin n° 114A et n° 114B sont conservés dans les registres du Trésor des chartes en JJ 80, n° 294, fol. 204r et JJ 91, n° 20, fol. 12r, puis le préambule n° 259 en JJ 91, n° 345, fol. 179v. Puis nous trouvons les préambules n° 32 et n° 73, qui étaient omis dans l’index, mais pour lesquels le corpus des préambules Barret-Grévin renvoie correctement aux registres de la chancellerie (JJ 105, n° 265, fol. 145r et JJ 105, n° 596, fol. 296v). Bref, au final, nous trouvons non pas seulement 17 préambules contenant « celsitudo » et recensés par Barret-Grévin, mais 22.

Si l’on compte d’après les occurrences repérées sur le nombre d’occurrences présentes, autant que l’on puisse savoir après nos repérages et mise à jour de la liste établie d’après l’index, il faut compter toutes les occurrences, qui se répartissent en : 18 occurrences non repérées ; 21 listées dans des sources autres que les registres du Trésor des chartes (en comptant Barret-Grévin n° 75), et 22 occurrences correctement identifiées (dont les deux omises dans l’index, mais présentes dans le corpus Barret-Grévin). À ce stade, le rappel est de 22/(22+18) = 55 %

Analyse des résultats quantitatifs

Une première remarque doit être faite quant à la précision et au rappel. Dans une comparaison entre les experts humains et la machine, si on compte les cinq occurrences supplémentaires, par rapport à une vérité-terrain « humaine », donc elle-même imparfaite, celle de l’index, on trouve 22 occurrences de « celsitudo » dans un préambule par rapport aux 35 de l’index, soit un rappel qui s’établit à 62,9 %. Évidemment, les Sciences de l’ingénieur ne peuvent pas réellement mesurer le rappel de cette façon, mais c’est une mesure réelle de ce que peut apporter la machine à nos recherches.

Une deuxième peut être formulée en comparant les 229 occurrences de notre recherche sur l’ensemble des registres numérisés actuellement aux 35 correspondant aux cotes JJ 80 – JJ 93 et JJ 105 qui sont listées dans le corpus Barret-Grévin, l’aide apportée aux recherches historiques est évidente. Des comparaisons plus approfondies seraient assurément nécessaires, mais l’étude du champ lexical et sémantique de « l’altesse royale » peut être entamée directement sans dépouillement systématique.

La troisième remarque vise les causes et mécanismes de la faute, et le développement de stratégies de recherche et d’indexation.

  • Plusieurs fois, l’origine de l’absence d’indexation est la mauvaise segmentation en ligne, parfois elle-même causée par des courbures sur l’image ou par la présence d’annotations marginales, comme l’analyse de l’acte – et le préambule est une partie particulièrement concernée, puisque l’analyse est en tête de l’acte.
  • Outre ces absences d’indexation, il faut aussi compter les erreurs de lecture et deux sont apparues de façon récurrente : « celsitudis » (abréviation non résolue) et « celsitudum » (erreur classique d’interprétation des jambages). La combinaison (celsitudis || celsitudum) apporte 28 résultats, et pas moins de quatre préambules pour dans les cotes JJ 80 – JJ 93 et JJ 105, ce qui améliore drastiquement le rappel et le fait monter à 65 %. Pour améliorer les résultats, tant de l’indexation que de la recherche, il est important de mieux implémenter les dictionnaires et outils de lemmatisation dans le moteur d’indexation, pour pousser la reconnaissance vers un terme existant, et, du côté de la recherche, créer une recherche floue qui pourrait intégrer des mécanismes de faute.

Dans un prochain billet, j’aurai le plaisir de faire quelques remarques sur les préambules dans une analyse historique.

Nota 1
ce texte a été rédigé le 28 septembre 2017, rendu visible sur le site le 24 novembre 2017.

Nota 2
Les données de l’évaluation sont rassemblées dans les différents onglets du tableur suivant :
HIMANIS : mesure de la précision et du rappel pour le mot “celsitudo” et comparaison avec le relevé Barret-Grévin

Job: Research Fellow position for the HORAE project / Ingénieur de recherche pour le projet HORAE

Job description / Fiche de poste (pdf) :

IRHT-HORAE-ResearchFellow-Ingénieur

I got a job for you… (Paris, Bibliothèque nationale de France, Dép. MSS, Rothschild 2535 (24 b)

English version: Research Fellow / Ingénieur de recherche
    Position focus
    Required Education, Skills and Experience
    About the project HORAE
    Host institution
    Salary
    How to Apply

Version française: Ingénieur de recherche en analyse de sources
    Missions
    Compétences
    Le projet HORAE
    Institution d’accueil
    Rémunération
    Candidature

 

Research Fellow / Ingénieur de recherche en analyse de sources

Institution: Institut de Recherche et d’Histoire des Textes

Location: Paris, France

Application Due: 31 Oct. 2017

Type: Full Time

Duration: 1 year, renewable until 3 years

Starting Date: between January 1, 2018 and April 1, 2018

 

The Institut de Recherche et d’Histoire des Textes (CNRS) invites applications for a Research Fellow (“Ingénieur(e) de recherché en analyse de sources”) for the project “HORAE – Hours, Analysis, Editions” (ANR-17-CE38-0008), under the supervision of Mr. Dominique Stutzmann, Principal Investigator.

Continue reading Job: Research Fellow position for the HORAE project / Ingénieur de recherche pour le projet HORAE

Saint-Omer in the registers of the French royal chancery (99.6% precision!)

As some of the readers will know, I am particularly interested in the city of Saint-Omer in Northern France. This is a good reason to test and search for the first time unknown occurrences of this city, its city officials and chapter officials in the registers of the French royal chancery.

As a matter of fact, the results are quite impressive…

Registrum factum in viagio sancti Audomari (Paris, Archives Nationales, JJ 88)

 

In total, 173 occurrences are found for “audomarum” OR “audomari” or “audomaro” and 73 occurrences for “saint omer”. I noticed bypassing that “osmer” is also present in the registers, but not in our index (generally silently regularized to “omer”).
Continue reading Saint-Omer in the registers of the French royal chancery (99.6% precision!)

“disant ses heures” : from HIMANIS to HORAE

We are delighted to announce that the French research agency is funding the HORAE project (ANR-17-CE38-0008). HIMANIS also highlights why it is important with one example:

Disant ses heures… (Paris, Archives Nationales, JJ44, fol. 98r)

un certain lieu ou il estoit alé disant ses heures

(translation: a place where he went saying his Hours; image Paris, Archives Nationales, JJ44, fol. 98r)

In another example, the book of hours of a Jehannette, daughter of Jehan Petit is central to the plot, since it is the cause of the crime! Used to force a date with the young Jehannette, it created a fight between the father and the daughter’s friend…

Thomas, par esbatement print les heures de ladite Jehannette qui estoient sur une fenestre dudit hostel en la presence de ladite Jehannette et de son consentement les emporta et dist a ycelle Jehannette qu’elle les alast querir a vespres en l’eglise, laquelle lui respondi que si feroit elle. Et tantost aprés, ledit Jehan Petit ala aprés ledit Thomas ou cymetiere de ladite eglise ou il estoit et de felon courage lui dist qu’il lui rendist ses heures, ou si non qu’il l’en poursuiroit et que faulsement et mauvaisement il les lui avoit emblees.

qu’il lui rendist ses heures (Paris, Archives Nationales, JJ 162)

At the core of the late medieval society…

HORAE (Hours – Recognition, Analysis, Editions) is a cross-disciplinary research project studying religious practices and experiences in the late Middle Ages through the books of hours, the absolute medieval best seller. In a public-private partnership, it gathers three partners in the Humanities and Computer Science: the Institut de Recherche et d’Histoire des Textes (IRHT, UPR 841), the private company TEKLIA and the Laboratoire des Sciences du Numérique de Nantes (LS2N, UMR6004). Books of hours forms, with more than 5000 preserved manuscripts, a vast and crucial ensemble to understand the medieval mindset. Yet, their textual content is very scarcely studied, although the massive production of such a large number of manuscripts is a pivotal cultural and industrial phenomenon to the profound changes in the medieval religious life. Not only a proto-industrial book production arises and the economics of demand shifts to the economics of offer, but also the awakening of the individual and the internalization of faith take places in an increasingly church controlled world. Books of hours are at the same time deluxe objects, implying social staging, and intimate objects, with devotional, memorial and eschatological challenges.

HORAE combines the research and expertise of all three partners in artificial intelligence applied to Computer Vision and image analysis, in Natural Language Processing (NLP), and in book history and religious practices. The research project aims at creating an integrated chain from image treatment to producing new knowledge by placing the end user at the center of the developments, by managing formats, interoperability and long term preservation, but also ergonomics and data visualization.The aims encompass: (1) re-using the many digitized manuscripts, which are online but underused resources; (2) new open source software for Handwritten Text Recognition (HTR); (3) tools for segmentation and plagiarism detection, adapted to the transcription of medieval manuscripts produced by the machine in order to identify the texts in the books of hours; (4) identifying and editing unpublished texts; (5) visualization of manuscript clusters, which share the same textual characteristics in the order of the different parts (Officium Beatae Mariae Virginis, votive offices, suffrages, prayers), but also in the order of textual units to identify the liturgical use; (6) studying the diffusion and circulation of devotional and liturgical texts at the end of the Middle Ages in order to better understand the cultures and faith in the 13th c.-16th c.

With its aims and methods, HORAE profoundly changes the ways of doing research in auxiliary sciences and tackles the challenges of big data. Books of hours have been too scarcely studied until new because they are too numerous, too complex and their text is very repetitive: now, their very number, repetitions and complexities make possible for this project to develop new and efficient technologies and methodologies and gain new knowledge on the Middle Ages.

 

 

Return on the language used for acts registration at the French Royal Chancery

 

The simple query of part of the protocol formula [ dei gratia francorum ] in the HIMANIS search engine return results that confirm what we already know about language used by the French Royal Chancery for recording acts.

http://prhlt-kws.prhlt.upv.es/himanis/index.php/ui/chapters/chancery?q=%5B+dei+gratia+francorum+%5D&t=50&feedback=1

Some warning.

  1. The complete formula ([name of the king] dei gratia Francorum rex…) is often skip and replace by “[name of the king] etc.”
  2. As the search engine is not yet lemmatized, distinction is made between “gratia” and “gracia”. To have more precise results, one can add the results of a second query [ dei gracia francorum ]
  3. Some noise might also be introduced by the vidimus copy of an act within an act.

Nevertheless, as from the first register (JJ35) the French Royal Chancery used both Latin and French and that one prevailed over the other at different periods, all major transitions are outlined by this query.

Serge Lusignan, in his 2004 book La langue des rois au Moyen Âge, point out how langage used evolved. From his observations, we can explain the major breaks-up in the following graph that we made from our query (number of query results are in red in the graph; in blue are the total number of acts for each register):

  1. JJ35-49: French is seldom used as language of registration by Philippe IV (5%)
  2. French is a little more used by Philippe IV two elder sons: JJ50-52 Louis X (19%) ; JJ53-60: Philippe V (10%).
  3. JJ53-65A: The French conquest begins slowly under his younger son, Charles IV (11% at the beginning; 26% at the end).
  4. JJ65B-79: Major break-up appears during the reign of Philippe VI. No reason is known, but in October 1330 French overthrow Latin for around 80% of act are registered in French from this date.
  5. JJ80-95: When Jean II become king (JJ80), Latin regain his place (2% of French only) until his capture. During the Regency of the Dolphin Charles (JJ84-95), French progressively take back his ground: from 32 % in 1358 it reaches 75% by the end of 1360. Back from his captivity, Latin is back again, but not with the past glory (21% French).
  6. JJ96-117: When Charles V becomes king in 1364 what today could be named a “Language policy” is put in place. From now, language is determined by the nature of the act recorded: French for remission; Latin for ennoblement, legitimation, Amortization and almost everything Church related. As Remission count for 62% of the registered act during his reign, French became predominant from now on.

99 pilgrimages / 99 “pelerinage(s)”: accuracy and historical research

Good news! We travelled across the registers, looking for pilgrimages… A search for pilgrimages in the registers of the French royal chancery is a fantastic example of what HIMANIS is bringing to the historians and interested citizens.

Search for “pelerinage” or “pelerinages” with a confidence level as low as 0 gives 54 matches with an average confidence of 46.6. This means 54 registers and 96 pages corresponding to the search, and 99 spotted occurrences. We searched in French only and, here, the results are amazingly accurate: All spotted occurrences are correct! And they even include an occurrence in a black-and-white register, scanned from the microfilm… A very good information for further research.

The table below shows the list of occurrences with direct links. Of course, there are some artefacts and in one occurrence, we see a bizarre result: http://prhlt-kws.prhlt.upv.es/himanis/index.php/ui/show/chancery/259/149?q=pelerinage+%7C%7C+pelerinages&feedback=1 in which there are two (red) boxes around the word, which is “pelerignage” but has also been indexed as “pelerinage”. In this case, I have to confess that I have clicked on “correct” to confirm the reading, before realizing that it was, technically speaking, not a correct match 😉
Continue reading 99 pilgrimages / 99 “pelerinage(s)”: accuracy and historical research

NEW : improved fonctionalities for correcting query results

We improve the way to correct the results by adding two new functionalities:

1/ After a query, on the spotted words, when you click on the word you are asked if the word is correctly spotted. Not entirely sure of the result you got, we add a third option (“not sure”)  to “correct” or “incorrect”.

2/ When clicking on a word to see if it has been recognized, you have now a dropdown menu suggestions of the recognized word in order of confidence. You can now just select the correct answer from the list rather then typing it. Of course, if none of the suggested result is good you can still type the correct word manually.

The bounding box surrounding the word now appears and they are in color (green, yellow, red or grey) so you can have an idea of the confidence level of the word you are going to select. And don’t worry if the suggest word in the bounding box have nothing to do with the one you clicked: the system will only take into account the pixel on which you clicked on.

Lemmatized search is on it’s way. We are still testing it on a small sample of the corpus but it will be implement in the coming month.

NEW : documents from BNF added to HIMANIS

New documents have just been added to the HIMANIS search engine. The new documents come from BNF digitized microfilm available on Gallica and that are related to the French royal chancery. Those are either originals or copies of chancery registers that are for various reasons located at the BNF rather then at the AN. Actual shelf marks of the chancery registers of the Trésor des Chartes come from the system put in place around 1380 by Gerard of Montaigu in the last version of his inventory of the libri utiles (actual JJ 115) who numbered the registers from I (JJ 11) to CXV (JJ 115). When the JJ serie was organized, between 1852 and 1854, numerous registers where added to the original serie that were not originaly in it:

  • Some came from what was considered by Montaigu as libi inutiles (JJ 42B, 79B, 267 à 270, 288, JJNN et JJOO).
  • Others came from the J serie (JJ 20, 24B -24D, 274 and 304-320) or even P (JJ 30B), Q1 (JJ 24A2) and X (JJ 54B) series.
  • Numerous are inventories and copies from the 13th to the 18th century (JJ 12A-JJ 128, JJ 9A, 9B, 27A-27C, 271-273, 275-303, 534-5884 et JJ A-JJ K).

In 1862 an exchange of documents took place between the AN and the BN and all of the above (except JJ 27C, 272 and 273) have been transferred to the BN in exchange for various inventories (nowadays JJ 12B-127).

Registers

– Français 5291 (ex JJ18): http://gallica.bnf.fr/ark:/12148/btv1b9059263p

– Français 10187 (ex JJ27B: http://gallica.bnf.fr/ark:/12148/btv1b90611100

– Français 18437 (ex JJ20): http://gallica.bnf.fr/ark:/12148/btv1b9063320p

– Latin 9778 (ex JJ27): http://gallica.bnf.fr/ark:/12148/btv1b90683044

– Latin 9784 (original of JJ47 and JJ277): http://gallica.bnf.fr/ark:/12148/btv1b9080793k

– Latin 9988 (final copy of JJ30A): http://gallica.bnf.fr/ark:/12148/btv1b9077723t

– Latin 9996 (inventory = JJ275 and 276): http://gallica.bnf.fr/ark:/12148/btv1b9076749r

– Latin 10919 (ex JJ29): http://gallica.bnf.fr/ark:/12148/btv1b9065926v

– Latin 10932 (ex JJ17): http://gallica.bnf.fr/ark:/12148/btv1b90807673

We also add the available chancery formularies and protocols.

Ptotocols and Formularies

– Latin 4763: http://gallica.bnf.fr/ark:/12148/btv1b9076661v

– Français 5024: http://gallica.bnf.fr/ark:/12148/btv1b9059518w (Formulary of Odart Morchesne, digital edition from ENC http://elec.enc.sorbonne.fr/morchesne/)

– Français 5030: http://gallica.bnf.fr/ark:/12148/btv1b90602939

– Français 5271: http://gallica.bnf.fr/ark:/12148/btv1b90604893

– Français 5318: http://gallica.bnf.fr/ark:/12148/btv1b9060505r

– Français 5727: http://gallica.bnf.fr/ark:/12148/btv1b9060472t

– Français 5909: http://gallica.bnf.fr/ark:/12148/btv1b9060762g

– Français 6022: http://gallica.bnf.fr/ark:/12148/btv1b9060611t

– Français 6142: http://gallica.bnf.fr/ark:/12148/btv1b9060599t

– Français 14371 (chancery protocol): http://gallica.bnf.fr/ark:/12148/btv1b9007185x

– Français 18674: http://gallica.bnf.fr/ark:/12148/btv1b9062281q

Accessing the full text of the Trésor des Chartes’ registers! (beta version)

HIMANIS project

Looking for the word “abbatisse” in plain text of the Tresor des Chartes’ registers

The HIMANIS project is funded by the Joint Programming Initiative on Cultural Heritage and Global Change” (JPI-CH) of the European Union. The partners are developing cost-effective solutions for querying large sets of handwritten document images. It gathers Computer Science, Humanities and Cultural Heritage institutions in order to produce technology to generate new, research-based knowledge from historical manuscripts. As a challenging and particularly interesting case study, the large collection of the Trésor des Chartes’ registers produced by the French royal chancery (Paris, Archives Nationales, JJ35 – JJ211).  The expected outcomes of this project are:

  1. a new indexing/searching technology for historical manuscripts
  2. a new paradigm to study our historical heritage, as conveyed by manuscripts, by using full text search technology.
  3. a new vision of the raise of nation states in Europe via a new study of the corpus under this paradigm.

Now we are proud to announce that we reach the two first goals of our project!  You can search the plain text in the Trésor des Chartes’ registers and provide feedback: It is ready to be used and tested by all interested users worldwide!

http://prhlt-kws.prhlt.upv.es/himanis/

You can search the plain text in the Trésor des Chartes’ registers and provide feedback!

How to search?

Query settings

Home page of the HIMANIS indexing and search engine
  • A query box is provided to type the queries
    • The search engine is case insensitive (“Paris” and “paris” retrieve the same results). The search engine allows for “AND”, “OR” and “EXPRESSION” queries (see below).
    • Abbreviations are not queried as they appear in the images; instead the corresponding expanded forms are used. For example, use “paris” to find “Paris”, “par.”, “par^s”, “Pars.”, etc.
    • Query text must be diacritcs-folded (e.g., “sebastien”, rather than “Sébastien”).
  • A confidence box and slider are provided to express a confidence threshold which determines the desired precision-recall tradeoff. The default is 50%. A higher value results in higher precision (little or none wrong spots) and lower recall (some, or many existing instances of the query may not be retrieved); a lower value results in lower precision (some, or many wrong spots) and higher recall (all or most the existing instances can be retrieved).
  • A ‘Max. results’ box is provided to express the maximum number of spots to be retrived and shown. For example, if this number is set to 20 and the confidence is set to 40%, the system shows at most 20 spots with confidence higher than 40%; and if no (or 0) confidence is set, the system retrieves at most 20 spots with the highest confidence.

    Query settings

Search: AND, OR, SEQUENCE

The search engine allows compound queries (boolean AND, OR, NOT) and “sequences”. For compound queries, the results are computed at Page level, and the number of matches is the total number of word occurrences matching the query.

  • Relations are interpreted at the full page-image level.
  • The operators are: AND: “&&” (or just blank space), OR: “||”, NOT: “-” (before the words which have to be negated). Parenthesis “(“, “)”, can be used for grouping
  • For “Sequence”, the number of match is computed at Page level and one match corresponds to the complete expression.
  • A query can be a sequence of words, expressed using the symbols “[” and “]”. Sequences are not exact strings; they allow (a few) extra words to appear among the stated words. Examples: “[ludovico francorum]” (where “francorum” is an expanded abbreviastion) would retrieve pages with strings such as “ludovico franc.”, “Ludovico rege franc.”, “ludovico dei gra. francorum”, etc. In the current system version, sequences can not be mixed with boolean operators.

Queries: NOT

“france – flandres”: pages with at least one instance of “france” but with no instance of “flandres”

Queries: mixed AND/OR/NOT with parentheses

“paris && (france || flandres)”: pages with at least one instance of “paris” AND at least one instance of either “france” OR “flandres”, or both

Navigation through the results

The current presentation of results is bound by the hierarchy of the collection and we know that it is not always easy to flip through the results. We ask for your patience and your help!

We ask for your patience and your help!

It is possible to:

  1. browse all registers
  2. browse all pages in one register
  3. see the list of registers with the number of matching pages (not the number of matches in the volume)
  4. see the list of pages with the number of matches

Queries are interpreted, and results are shown, hierarchically. For example, assume a collection is composed of many Volumes and each Volume of many Pages. Then, if the query is issued at the top collection level, the system first shows the Volumes where the query may appear with confidence above the threshold specified. Then if the same query is issued for a specific Volume, the system shows all the Pages where the query may appear with a confidence above the threshold. Finally if it is issued for a specific Page, the system shows locations where the words involved in the query may appear. In the PRHLT interface, default levels of the hierarchy are called: “Home”, “Collection”, “Chapter” and “Page”.

The navigation from one page goes either to the previous or the next one in the same volume, or back to a higher level (volume or corpus). There is no possibility to jump directly to the next match.

To go back to the superior level, please navigate with the breadcrumb.

Number of volumes with at least one match (and average confidence)
List of volumes with at least one match and number of pages with at least one match
List of pages in the volume and number of matches on the page
Matches on the page
  1. Miniatures. At each level of the hierarchy (except at the lowest one, the page level), the system shows the elements of this level by means of miniature images, with some information associated; namely,
    1. an identifier of the element,
    2. a bar representing the confidence of the system that the query appears somewhere in this element
    3. the number of elements in the lower level where the query may appear with a confidence above the threshold specified.
  2. By hovering the mouse over the confidence bar, the actual confidence value is shown.
  3. Pan and Zoom: When a (part of a) page image is shown, it can be explored by moving the mouse while holding its right button. In addition, the mouse wheel can be used to zoom in and out.
  4. If a word appears more than once in a text line, only the instance with greatest confidence is shown.
  5. The approximate locations of spotted queries in each page are marked with rectangles (called “bounding boxes”) surrounding the corresponding words. The color of a rectangle expresses the degree of confidence the system has in the corresponding spot. Exact confidence values can be seen by hovering the mouse over the rectangles.

Your feedback is needed!

YOU CAN GIVE YOUR FEEDBACK

On the spotted words

When you see the results of a query a “word level”, you can click on the word and you are asked if the word is correctly spotted. Please help us and say “correct” or “incorrect”.

Giving feedback on spotted words

If you see a missed word

In the example below, the word “pedagiis” has been correctly spotted on the lower right part of the image.

Correct and missed matches on one page

If you double-click on the image, you can add the word to our index!

Inserting the correct transcription for a word

Hints for the Chancery collection

  • Click on the Chancery miniature with no query to see the 147 Volumes indexed, each showing the number of images it contains (ranging from less then 100 to more than 1000).
  • Queries can be formulated at any level: full Collection (Chancery), specific Chapter (Volume JJ…), or individual Page.
  • Remember to use the Confidence and/or Max.results levers depending on the precision-recall tradeoff you wish to achieve.
  • The number of the “page” indicated is the number of the file.
  • There are 147 volumes; 32 additionnal volumes are being digitized by the French National Archive

In the near future…

This is a presentation of the Beta Version. In a near future, you will be able to scroll through the results and provide feedback directly on a hit list of the results.

You will be able to get the correct foliation of each image and retrieve metadata from the inventories and text editions.

IRHT’s interface proposal to access the Chancery corpus: aligning and merging early modern inventories (tables of contents), 20th c. inventories and indexes, partial editions, and the result of automated indexing performed in the HIMANIS project

A glimpse at the coverage of extant inventories

Last week, I was in Graz, talking about “Automated indexing of medieval manuscripts: the HIMANIS research project and the registers of the French royal chancery” at a Lunchtime Lecture of the Zentrum für Informationsmodellierung at the University of Graz.

For this lecture, I prepared a small slide which I had ever wanted to do: just give an idea of the coverage of the inventories at a glimpse. Well, it is easy and difficult at the same time.

The Virtual Reference Room of the National Archive indicates many inventories for the fonds JJ. Not fewer than 42!

But, if you simplify a bit, then, you have the systematic printed inventories from JJ 37 to JJ 79 (1300-1350), then you have handwritten unpublished inventories for the years 1350-1364 (JJ 80 to JJ 95), and, besides some particular text editions, esp. regarding the Poitou region, some geographic inventories for JJ 35 to JJ 266, covering the provinces of Gascogne, Blésois, Orléanais, Chartrain, Languedoc and Rouergue.
In the map above, there are of course some catches: I’ve avoided the question of the exact limits of the older provinces by using a map with actual departements. But I hope this gives a good idea of how much we have to discover in the Chancery registers 😀

TEI as a format for archive inventories? At which level is the “text”?

As mentioned in previous posts, we are dealing with plenty of inventories in the HIMANIS project: handwritten systematic inventories, printed ones, cardboard files, geographic inventories, indexes, etc.

Now, the issue we have, is of the coherence of the formats we are using. The text edition we use as a training data set and ground truth, namely the Actes royaux du Poitou edited by Paul Guérin and converted by the Ecole nationales des Chartes, is in TEI-P5 format. There are twelve independent TEI files. This is normal because it is a text edition.

But what would you do with inventories? Some are plain text, some are in EAD format. We could convert them to TEI and consider that these inventories are, each of them, one “text”, so that we could convert each of them to a single TEI file, using the <body> to render the core of the inventory.

For the systematic inventories, we also could consider that each of the inventories contains “manuscript descriptions”: each register can be rendered within a <msDesc> and each text is a <msItem> within the <msContents> part.

The volumes which are copied the one from the other (such as JJ 42B and JJ41) makes it difficult to keep the same structure, except if you want to multiply the descriptions, with one <msDesc> for each of the registers and each <msItem> having the same content but the respective foliation.

In HIMANIS, we are making a new choice: considering each volume as a <TEI> entity (combining them thanks to <teiCorpus>), whose text gathers the <group> of all contained royal charters, each of which being a <text>. The abstract and analysis remains in the <front> part, with an empty <body>.


This solution is, obviously, borderline, since we end up having no text in the whole TEI document, but it is an opening towards a future in which the text could be added as if it where a scholarly digital edition. On the top of that, of course, it allows us to homogenize all source formats, including the text editions, at the (almost) the same level towards an integrated environment. (Not exactly the same level, since the <TEI> file for Guérin is one volume of the edition, not one volume of the register).
For all comments of the community, we are grateful!

Update on the inventories of the Chancery registers

Inventory of Archives Nationales, JJ 84 (11v-12r)

On December 5, during our second plenar meeting, held in Valencia (Spain), we discussed (among other topics) various issues that we have encountered with the inventories of the French royal chancery registers. Until the HIMANIS project, those were the only tools available to find particular information in those registers. They exist in different formats: printed or handwritten; some are already converted in EAD (in different flavours), but, despite the efforts of the Archives Nationales, not all available for the general public yet.

As we intend to use these inventories in the project in order to collect correct identifications for place and person names, as well as to evaluate the precision/recall of indexing, we need to have them in a interoperable electronic format (TEI / EAD).

The oldest inventory which we will use, and the only one covering the entire corpus, is the three volume manuscript inventory made in the nineteenth century by A. Longnon and A. Coulon, providing very short Latin summaries for each document. This inventory was retyped by the Archives Nationales and converted in the EAD format.

A second series of inventories encompass the geographical inventories which were elaborated for Gascogne; Languedoc and Rouergue; Berry, Blésois, Chartrain, Orléanais, Touraine. They cover the whole corpus (JJ 35 – JJ 266), but not all the acts in the corpus as they are limited by the region that they are interested in. It is also a short summary with index. We converted the OCR made by Gallica and converted them into TEI as already mentioned a year ago in our post on Named entities and ground-truth: progress report

Modern systematic inventories elaborated by the Archives Nationales come in different formats: printed, typed or are still at manuscript stage (Registres du Trésor des Chartes, Inventaires analytique, 6 vol. in 3 t.).

  • T. I: Philippe IV (1300-1314) [JJ 37 – 50]
    • OCR by AN; EAD with subjects index coded
    • Index (place names + persons) : TEI by IRHT
  • T. II: Fils de Philippe le Bel (1314-1328)
    • T.II.1 : Louis X, Philippe V (1314-1321) [JJ 50 – 60]: EAD by AN
    • T.II.2 : Charles IV (1322-1328) [JJ 61 – 64]: TEI by IRHT
    • No index existing
  • T. III: Philippe de Valois (1328-1349)
    • T. III.1 1328-1349 [JJ 65A – 69]: OCR by IRHT; TEI
    • T. III.2 1336-1346 [JJ 70 – 75]: OCR by AN; EAD
    • T. III.3 1339-1350 [JJ 76 – 79B]: OCR by AN; EAD
    • Index (subjects, person, place names, chancery notes): OCR by IRHT; TEI

Typed unpublished inventories have been retyped by the Archives Nationales (AN) :

  • Jean II and Charles Dauphin
    • JJ 80 – JJ 81: OCR by AN; plain text converted to TEI
    • JJ 83: OCR by AN; .xls to be converted to TEI
    • JJ 82 – 95: Ms. Index on cardboard files (persons and subjects) about 30.000 files; re-typed by AN
    • JJ 96 – 97: Ms. Index on cardboard files (persons, place names, subject): 4.500 files; retyped by AN
  • Charles V (1364-1366)rc
    • JJ 96 to JJ 98: abstracts and index in .xls to be converted to TEI

Last but not least, our main concern were with the 12 inventories that  were still in handwritten stage (JJ84-95 or about 1400 pages). Those where remained in the manuscript (JJ84-95 for the reign of Jean II and a part of Charles V). We therefore asked our partner A2ia if they can perform HTR on those. But how did we proceed to achieve this task?

 

First, the AN provided us with high resolution reproduction of those 12 inventories, or 1429 pages, at 300 dpi in tiff format. We then produce ground truth, which consisted of the 4 first pages of each inventories (1079 lines of text). Then A2ia trained and performed HTR. We were lucky as the inventory have only one hand and as the layout is very stable from one inventories to the other. Using a standard French model and lexicon from the printed inventories, this operation was able to be carried out quickly in less then 4 weeks and produced results that went beyond our expectations with a very high recognition rate and transcript almost usable as is.

Inserted acts and vidimus copies in the registers of the French royal Chancery: Originals without originality?

During the colloquium L’acte original, entre conceptions médiévales et concept diplomatique (Nancy, 6-7 oct. 2016), we presented a paper about the practice of “insertion”, which consists in transcribing an earlier document into the body of another act. This practice is used, according to the Vocabulaire international de diplomatique (International vocabulary of diplomatics), either to give legal support to this act (for example, by transcribing the powers of prosecutors or commissions), or to confer authenticity to the transcribed act, known as “vidimus”. This double practice is massive in the registers of the French Royal Chancery, going as far as to produce vidimus of vidimus.

Using the materials accumulated so far in the HIMANIS project, encompassing several text editions and all available inventories (despite the differences in structure),  we quantified this practice and documented its evolution from the reign of Philip IV (1286-1314) until Louis XI (1461-1483) and were able to show diachronic variation (with a maximum under the reigns of Philip IV’s sons and a sharp decrease under Charles V and his successors) as well as geographic differences.

We also studied the diplomatic features (beneficiaries etc.) and indications of collation (becoming quasi systematic under Philip VI) and mise en texte within the registers.

HIstorical MANuscript Indexing for user-controlled Search