====== GBX9MO23 - Information Access and Retrieval 2021-2022 ======

<!-- ** Zoom link for next course (January 8, 2021): https://univ-grenoble-alpes-fr.zoom.us/j/91550419498?pwd=c3BGaWh3NlZLQStpWEltNHJnOFJLQT09 ** -->

<!--\\
\\
==== February 1st, 2021, 9-11am examination ====

2020-2021 examination: {{ :gbx9mo23-2020-2021-exam-tmpkxtz.pdf |}}
\\
\\
\\
\\-->

Available from Master 2 [[http://mosig.imag.fr/ProgramEn/M2S1|MOSIG]]
and [[https://iam.imag.fr/m2tracks#data_science_ds|MSIAM]].
[[http://formations.univ-grenoble-alpes.fr/en/catalog/master-s-degree-XB/sciences-technologies-and-health-STS/master-in-computer-science-program-program1-master-informatique-en/master-of-science-in-informatics-at-grenoble-mosig-subprogram-subprogram-master-of-science-in-informatics-at-grenoble-mosig-en/ue-information-access-and-retrieval-IGDF9Z8M.html|Course Description]]


This course is given by
[[http://mrim.imag.fr/User/jean-pierre.chevallet/|Jean-Pierre Chevallet]],
[[http://lig-membres.imag.fr/mulhem/|Philippe Mulhem]],
[[http://mrim.imag.fr/User/lorraine.goeuriot/|Lorraine Goeuriot]] and
[[http://lig-membres.imag.fr/quenot/|Georges Quénot]] from the [[http://lig-mrim.imag.fr/|Multimedia Information Modeling and Retrieval]] (MRIM) research group of the [[https://www.liglab.fr/en|Grenoble Informatics Laboratory]] (LIG).

Contact: [[mailto:Georges.Quenot@imag.fr"|Georges.Quenot@imag.fr]]

Contents / schedule<!-- (might be updated)-->:

Part I. Foundations of Information Retrieval (Jean-Pierre Chevallet and Philippe Mulhem)
  * Course 1: {{m2r_mosig_ira_chapter_01_information_retrieval_basics.pdf|Information retrieval basics}} (J.-P. Chevallet).
  * Course 2: {{m2r_mosig_iar_chapter_02_classical_models_for_information_retrieval.pdf|Classical models for information retrieval}} (J.-P. Chevallet).
  * Course 3: Vector space model and embeddings (J.-P. Chevallet)
  * Course 4: {{ :probabilistic_information_retrieval_2020.pdf |Probabilistic IR Models (version for 2020)}} (P. Mulhem).{{ ::exercices_lm_-_correction.pdf | 2 exercices with corrections}}
<!--  * Course 4: {{m2r_mosig_iar_natural_language_processing_for_information_retrieval.pdf|Natural language processing for information retrieval}} **(NB: not presented this year)** (J.-P. Chevallet).-->

Part II: Web, social networks and health (Philippe Mulhem, Lorraine Goeuriot)
  * Course 5: {{ ::web_ir_2021.pdf |Web information retrieval (NEW: 2021 version)}} (P. Mulhem). {{ ::hits-pr.xlsx | Excel document with the computations of Hits and Pagerank values of the slides}}. {{ ::log_regression_ex.xlsx | Excel document with the example of logistic regression for Learning to Rank of the slides}}.
  * Course 6: {{ ::evaluation_of_information_retrieval_-_2021.pdf | IR evaluation}} (2021 version, <color #ed1c24>last update 23rd of November 2021</color>) (P. Mulhem). {{ :computing_eval_examples.xlsx |Excel document with examples of Recall/precision tables and diagrams, and nDCG}}.<!--  8-o The video recording of the lesson is on INP Chamillo, under  ENSIMAG WMM9MO23 Information access and retrieval/Documents/Course6_IR_Evaluation.mp4 .-->
  * Course 7: Following of course 6.
<!-- {{ ::personnalization-social_-_2020.pdf |Personalization and social information retrieval (2020 version)}} (P. Mulhem).--><!-- 8-o The video recording of the lesson is on INP Chamillo, under  ENSIMAG WMM9MO23 Information access and retrieval/Documents/Course8_IR_Personalization_Social_IR.mp4 .-->
  * Course 8: {{::medical-ir.pdf|Medical information retrieval}} (L. Goeuriot).


Part III: Multimedia indexing and retrieval (Georges Quénot)
  * Mathematics Reminders: {{:mr1.pdf |Linear algebra and convolutions}} (G. Quénot).
  * Mathematics Reminders: {{:mr2.pdf |Differential calculus}} (G. Quénot).
  * Course 9: {{:m2-mosig-iar-9.pdf|Visual content representation and retrieval}} (only introduction and color / texture / points of interest descriptors) (G. Quénot).
  * Course 10: {{:m2-mosig-iar-10.pdf|Classical machine Learning for multimedia indexing}} (part on LSH is excluded) (G. Quénot).
  * Course 11: {{:m2-mosig-iar-11.pdf|Deep learning for multimedia indexing and retrieval - part 1}} (G. Quénot).
  * Course 12: {{:m2-mosig-iar-12.pdf|Deep learning for multimedia indexing and retrieval - part 2}} (G. Quénot).

<!--8-o **Video recodings**

You can find the videos of courses 5, 6, 8, 9 and 10 on INP Chamillo, under  ENSIMAG WMM9MO23 Information access and retrieval/Documents. \\
You can get the videos of courses 11 and 12 from the MRIM server at: \\
http://mrim.imag.fr/georges.quenot/gbx9mo23/Course_11_Deep_learning_for_multimedia_indexing_and_retrieval_part_1.mp4 \\
http://mrim.imag.fr/georges.quenot/gbx9mo23/Course_12_Deep_learning_for_multimedia_indexing_and_retrieval_part_2.mp4 \\
The other courses were not recorded. -->

<!-- **Warning:** the course contents currently online are those of last year, all of them are expected to be updated, possibly significantly for some of them. We recommend that you do not print them for now. -->


Reference to IR books or papers
  * [[http://nlp.stanford.edu/IR-book/|Introduction to Information Retrieval, http://nlp.stanford.edu/IR-book/]]
  * [[https://ciir.cs.umass.edu/irbook/|Search Engines Information Retrieval in Practice, https://ciir.cs.umass.edu/irbook/]]

===== Practical =====

The goal of this practical is to implement in Python a skeleton of an IR system.
The fundamental structure to implement is a inverted file that contents for each term,
a document internal Id (integer) and the frequency (integer) of this term in this document. 

You must accept the following implementation constraints to reduce the size of this inverted file:
  * Document Id stored on a 4 bytes unsigned integer (32 bits, max 4 millions of indexed documents)
  * Frequency stored on only 1 byte unsigned integer (8 bits, max frequency of 256)
For that, use the python basic and efficient [[https://docs.python.org/3/library/array.html|array]] structure.
This is the only imposed data structure for this project.

The inverted file can be loaded in memory for more efficiency when searching.
Beside that, you have to store document external Id in a sequential structure that can be saved.
Also, the dictionary should be programmed using a basic python structure, that can be saved.

Optionally, you can used the following libraries:
  * The [[https://www.nltk.org/api/nltk.corpus.html|nltk.corpus]] part of the Natural Language Toolkit for reading the corpus and for the stopwords
  * The [[https://www.nltk.org/api/nltk.stem.html|nltk.stem]] for stemmers like [[https://www.nltk.org/api/nltk.stem.snowball.html|snowball]].


For the data set, you can use the following test collection (document and solved query) :
  * The data ressources available from [[http://ir.dcs.gla.ac.uk/resources/|Glasgow University IR Research Group]] like:
    * [[http://ir.dcs.gla.ac.uk/resources/test_collections/cacm/|CACM]]
    * [[http://ir.dcs.gla.ac.uk/resources/test_collections/cran/|Cranfield]]
    * ...
  * Or other ressources like [[http://www.daviddlewis.com/resources/testcollections/reuters21578/|reuters21578]]



The minimum matching model to implement are the Vector Space Model and one simple Language Model.

You can do you projet up to 3 persons and you have to send the following elements packaged in a compressed file:
  * The full source code in Python (without the cache .pyc files to reduce size)
  * A minimal documentation with the name of each member of the group and explanation on how to use your system and of on witch test collection it works, witch IR model, etc.

Please do not include any data to reduce the file final size.
Produce a minimalist short code: the goal of this project is to better understand IR system real working.  

Then send the result to : jean-pierre.chevallet@univ-grenoble-alpes.fr, before the official exam period (around end of January 2022.
Meanwhile, you can send technical question to jean-pierre.chevallet@univ-grenoble-alpes.fr.



===== First session examination =====

The examination will be on **February 2nd, 2022 from 9:45 to 11:45am** in ENSIMAG **Amphi E**. \\
Course materials, the two papers related to the examinations, personal notes, and calculators (without network capabilities) are allowed. 


You are expected to do a research work on the two papers proposed below, in a way to understand them and to be able to comment then.
You will have to answer questions on topics that occur in the lessons. 
You must also take time to read complementary information in order to understand the papers. 
**Be sure to bring with you a copy of the two research papers as they will NOT be redistributed with the examination subject.** These can be annotated by you. The bibliography and appendices, if any, are part of the papers.

<color #ed1c24>The papers for the 2021/2022 exam are:</color>
  * B. Mansouri, R. Zanibbi, D. W. Oard, Learning to Rank for Mathematical Formula Retrieval, ACM SIGIR '21, https://www.cs.rit.edu/~rlaz/files/LTR_Formulas_SIGIR2021.pdf
  * Jiaxin Wu and Chong-Wah Ngo, Interpretable Embedding for Ad-Hoc Video Search, ACM Multimedia 2020, http://vireo.cs.cityu.edu.hk/papers/MM2020_dual_task_video_retrieval.pdf

<!--  * F. Faghri, D. J. Fleet, J. R. Kiros, S. Fidler, \\ VSE++: Improving Visual-Semantic Embeddings with Hard Negatives, \\ Proceedings of the British Machine Vision Conference (BMVC), 2018. (BMVC Spotlight) \\ https://arxiv.org/pdf/1707.05612
  * Zamani, H., Dehghani, M., Croft, W. B., Learned-Miller, E., & Kamps, J. (2018, October). \\ From neural re-ranking to neural ranking: Learning a sparse representation for inverted indexing. \\ Proceedings of the 27th ACM international conference on information and knowledge management (pp. 497-506).\\ https://people.cs.umass.edu/~elm/papers/zamani.pdf -->





==== Previous years examinations ====

2017-2018 examination: {{ :gbx9mo23-2017-2018-exam.pdf |}}, papers: \\
https://www.researchgate.net/publication/305081616_A_Simple_Enhancement_for_Ad-hoc_Information_Retrieval_via_Topic_Modelling, \\
[[http://www.tyr.unlu.edu.ar/tallerIR/2014/papers/novel-tfidf.pdf]], \\
[[http://openaccess.thecvf.com/content_cvpr_2017/papers/Huang_Densely_Connected_Convolutional_CVPR_2017_paper.pdf]]

2018-2019 examination: {{ :gbx9mo23-2018-2019-exam.pdf |}}, papers: \\
[[https://danluu.com/bitfunnel-sigir.pdf]], \\
[[https://arxiv.org/pdf/1604.01325]].

2019-2020 examination: {{ :gbx9mo23-2019-2020-exam.pdf |}}, papers: \\
[[http://openaccess.thecvf.com/content_CVPR_2019/papers/Dong_Dual_Encoding_for_Zero-Example_Video_Retrieval_CVPR_2019_paper.pdf]] \\
[[https://ciir-publications.cs.umass.edu/pub/web/getpdf.php?id=1302]]

2020-2021 examination: {{ :gbx9mo23-2020-2021-exam-tmpkxtz.pdf |}}, papers: \\
[[https://arxiv.org/pdf/1707.05612]] \\
[[https://people.cs.umass.edu/~elm/papers/zamani.pdf]]

<!-- ===== Second session examination =====

The second session examination will take place **on April 16, from 8am to 10am**. It will be an oral examination and it will he held remotely via zoom with these links:
  * Jean-Pierre Chevallet: https://univ-grenoble-alpes-fr.zoom.us/my/chevallj
  * Georges Quénot: https://univ-grenoble-alpes-fr.zoom.us/my/quenotg
You will individually receive an email with the exact time at which you should reach these rooms. You will have 15 minutes with Georges Quénot and 15 minutes with Jean-Pierre Chevallet. There is no new paper to study for this session. Questions and/or exercises will be related to the papers selected for session 1 (above) with some related to the course in relation with the papers. -->

<!-- ===== Internship projects proposed by the MRIM team (more to come) =====

[[http://im2ag-pcarre.e.ujf-grenoble.fr/CDExportProjetHtml?idSujet=2868|Explainable AI: Improving local model explanations in visual recognition tasks]] Georges Quénot (LIG-MRIM) and Marie-Christine Rousset (LIG - SLIDE). -->

<!-- [[2018-2019 page]] -->
<!-- [[2019-2020 page]] -->
<!-- [[2020-2021 page]] -->
[[2021-2022 page]]